Learning my per-matchup MMR in Starcraft II through PyMC3
In this post we'll continue our SC2 replay research, started last time. You may want to go back to that and pick up on the terminology!
To recap: we used replay data from my SC2 games over 2019 to estimate a "true MMR" value and infer the size of per-game fluctuations. This time, we'll redo that analysis, except to get something more useful: we'll look at the three matchups I played and infer separate MMR values for each of those. Let's dig into it!
I'll redo the basic data cleaning steps here. If any of this is confusing, reviewing the previous post might really be a good idea - or you could ask a question below, as always!
If you take a close look, you might also find a teaser for one of the next posts in this series here :)
import pandas as pd
import altair
def MMR_winrate(diff):
return 1 / (1 + 10**(-diff/880))
df = pd.read_csv("https://raw.githubusercontent.com/StanczakDominik/stanczakdominik.github.io/src/files/replays.csv", index_col=0)
df['time_played_at'] = pd.to_datetime(df.time_played_at)
df = df.sort_values('time_played_at')
for column in ['race', 'enemy_race', 'map_name']:
df[column] = pd.Categorical(df[column])
df['enemy_mmr'] = df['mmr'] - df['mmr_diff']
df['expected_winrate'] = MMR_winrate(df.mmr_diff)
all_data = df[(df.mmr > 0) & (df.enemy_mmr > 0) & (df.race == "Protoss") & (df.duration > 10)]
all_data = all_data.rename({"enemy_nickame": "enemy_nickname"}, axis=1) # whoops
data = all_data[(all_data['time_played_at'] > '2019-01-01') & (all_data['time_played_at'] < '2020-01-01')]
data
time_played_at | win | race | enemy_race | mmr | mmr_diff | enemy_nickname | map_name | duration | enemy_mmr | expected_winrate | |
---|---|---|---|---|---|---|---|---|---|---|---|
8 | 2019-10-06 12:36:36+00:00 | True | Protoss | Protoss | 3826 | 78 | vasea | World of Sleepers LE | 743 | 3748 | 0.550847 |
325 | 2019-10-08 19:33:28+00:00 | False | Protoss | Protoss | 3893 | -53 | Wavelength | Ephemeron LE | 254 | 3946 | 0.465386 |
54 | 2019-10-10 07:41:27+00:00 | False | Protoss | Zerg | 3828 | 26 | PereiRa | Winter's Gate LE | 45 | 3802 | 0.517001 |
346 | 2019-10-10 07:55:19+00:00 | True | Protoss | Zerg | 3760 | -56 | <PROOO><sp/>Jesperpro | Thunderbird LE | 801 | 3816 | 0.463433 |
138 | 2019-10-10 20:42:11+00:00 | True | Protoss | Protoss | 3827 | 126 | Pippuri | Acropolis LE | 697 | 3701 | 0.581684 |
... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... |
391 | 2019-12-27 20:24:27+00:00 | False | Protoss | Zerg | 3933 | -100 | HiveMind | World of Sleepers LE | 262 | 4033 | 0.434956 |
25 | 2019-12-27 20:40:39+00:00 | True | Protoss | Zerg | 3914 | 0 | Racin | Nightshade LE | 911 | 3914 | 0.500000 |
208 | 2019-12-27 21:24:06+00:00 | True | Protoss | Terran | 3936 | -41 | <DemuCl><sp/>Jazzz | Nightshade LE | 1277 | 3977 | 0.473206 |
59 | 2019-12-28 20:58:25+00:00 | True | Protoss | Terran | 3959 | 22 | rOoSter | Simulacrum LE | 76 | 3937 | 0.514387 |
364 | 2019-12-28 21:06:48+00:00 | True | Protoss | Zerg | 3980 | -260 | contremaitre | Nightshade LE | 478 | 4240 | 0.336192 |
138 rows × 11 columns
Let's visualize the games on a per-matchup MMR vs enemy MMR basis. I added some fancy Altair selection magic, so you can look at winrates in specific MMR ranges.
brush = altair.selection(type='interval')
scatter = altair.Chart(data).mark_circle().encode(
altair.X('enemy_mmr',
scale=altair.Scale(zero=False)),
altair.Y('mmr',
scale=altair.Scale(zero=False)),
facet='enemy_race',
size='expected_winrate',
color='win',
tooltip='enemy_nickname',
).add_selection(brush)
bar = altair.Chart(data).mark_bar().encode(
x=altair.X('mean(win):Q', scale=altair.Scale(domain=(0, 1))),
y='enemy_race:O',
).transform_filter(brush)
scatter & bar
Separate matchup MMRs¶
This is where the magic starts. Where, before, we had a single MMR estimation, we'll now have three, one for each matchup: $$\mu^n \sim \text{Normal}(4000, 300) \text{ for } n \text{ in } \{1, 2, 3\}$$
And likewise for the fluctuation value: $$\sigma^n \sim \text{HalfNormal}(100)$$
And that, honestly, is about it! When I realized it, I wanted to title this post "How Can It Be That Simple, Like, What The Hell". But I did have to tinker with the model for a good while to find out the optimal way of doing things. It turns out the first idea I had was optimal. Who knew.
We'll use some fancy new PyMC3 3.9 and ArviZ 0.8.3 functionality to replace the old shapes
arguments with dims
, for cleaner code.
Note: to reproduce, use the GitHub master release of ArviZ for now.
import pymc3 as pm
import arviz as az
# fancy new functionality for xarray output - I'll explain later!
coords = {
"replay": data.index,
"race": ["Terran", "Protoss", "Zerg"],
}
We now assign the new priors for $\mu^n$ and $\sigma^n$, three of each - and then we'll add a helper variable for each of the replays. Note how the new syntax is a good bit cleaner than hardcoding the shapes in.
with pm.Model(coords=coords) as split_model:
mmr_μ_matchup = pm.Normal('μ', 4000, 300, dims='race')
mmr_σ_matchup = pm.HalfNormal('σ', 100, dims='race')
mmr_σ_norm = pm.Normal('helper', 0, 1, dims='replay')
And the next change we have to make is indexing the per-race average and fluctuation values based on the enemy races, so that each game in our dataset gets the MMR for its particular matchup.
We'll have to assign a numerical index for each possible enemy race. We'll choose zeroes for Terran and two for Zerg, so that, at least in indices, Protoss can be number one.
race_encoding ={"Terran": 0,
"Protoss": 1,
"Zerg": 2}
with split_model:
enemy_races = pm.Data("enemy_race", data.enemy_race.map(race_encoding).astype(int), dims='replay')
mmr = pm.Deterministic('MMR', mmr_μ_matchup[enemy_races] + mmr_σ_matchup[enemy_races] * mmr_σ_norm, dims='replay')
And now it's smooth sailing from here on out! I forgot to add it last time, but PyMC3 can create a neat graph for your model using GraphViz.
If I had remembered to do so, the only difference between this model and ours would be the 3 for $\mu, \sigma$ - since we now have three of each - and adding the enemy_race
as pymc3.Data
.
with split_model:
enemy_mmr = pm.Data("enemy_mmr", data.enemy_mmr, dims='replay')
diffs = pm.Deterministic('MMR_diff', mmr - enemy_mmr, dims = 'replay')
p = pm.Deterministic('winrate', MMR_winrate(diffs), dims = 'replay')
wl = pm.Bernoulli('win', p=p, observed=data.win, dims = 'replay')
pm.model_to_graphviz(split_model)
First run¶
And now, let's sample
! We'll add a predictive prior and posterior sample: this lets us easily see what sort of data we'd see from our initial assumptions and from the fully "learned" ("taught"?) model.
predictive_var_names = "win μ σ winrate".split()
with split_model:
trace = pm.sample(2000, tune=2000, chains=4, random_seed=1)
output = az.from_pymc3(trace=trace,
prior=pm.sample_prior_predictive(2000 , var_names=predictive_var_names, random_seed=1),
posterior_predictive=pm.sample_posterior_predictive(trace, var_names=predictive_var_names, random_seed=1),
)
output
Auto-assigning NUTS sampler... Initializing NUTS using jitter+adapt_diag... Multiprocess sampling (4 chains in 2 jobs) NUTS: [helper, σ, μ]
Sampling 4 chains for 2_000 tune and 2_000 draw iterations (8_000 + 8_000 draws total) took 18 seconds. There were 4 divergences after tuning. Increase `target_accept` or reparameterize. There were 3 divergences after tuning. Increase `target_accept` or reparameterize.
-
-
- chain: 4
- draw: 2000
- race: 3
- replay: 138
-
-
chain(chain)int640 1 2 3
array([0, 1, 2, 3])
-
draw(draw)int640 1 2 3 4 ... 1996 1997 1998 1999
array([ 0, 1, 2, ..., 1997, 1998, 1999])
-
race(race)<U7'Terran' 'Protoss' 'Zerg'
array(['Terran', 'Protoss', 'Zerg'], dtype='<U7')
-
replay(replay)int648 325 54 346 138 ... 25 208 59 364
array([ 8, 325, 54, 346, 138, 405, 20, 129, 104, 46, 302, 408, 101, 219, 316, 126, 98, 231, 241, 385, 193, 198, 90, 329, 137, 200, 80, 355, 317, 33, 64, 213, 368, 49, 435, 134, 254, 330, 60, 39, 218, 109, 301, 133, 328, 181, 156, 395, 43, 249, 27, 153, 211, 420, 366, 186, 163, 63, 202, 45, 69, 31, 167, 177, 95, 151, 392, 387, 18, 286, 102, 290, 195, 428, 403, 97, 406, 412, 374, 263, 371, 41, 212, 52, 238, 345, 4, 117, 407, 56, 103, 118, 319, 57, 128, 294, 15, 273, 327, 281, 378, 121, 113, 284, 422, 389, 216, 58, 418, 309, 123, 116, 222, 122, 11, 361, 179, 255, 239, 225, 381, 424, 343, 287, 341, 184, 380, 196, 174, 306, 252, 148, 416, 391, 25, 208, 59, 364])
-
-
-
μ(chain, draw, race)float644.303e+03 3.375e+03 ... 3.955e+03
array([[[4302.79553938, 3375.01011989, 4048.86315149], [4323.86457361, 3295.74222841, 3954.05048638], [4239.6739907 , 3448.28894514, 4235.71022401], ..., [4250.95500899, 3592.96680679, 3907.78214913], [4321.36051851, 3657.60296464, 4041.53244852], [4378.58347689, 3680.20202035, 4041.44904971]], [[4141.06331072, 3385.94471985, 4135.36324411], [4138.80737046, 3670.4073407 , 4045.58916728], [4194.72695819, 3676.40161761, 3931.11602939], ..., [4066.10358294, 3429.85753425, 4004.41005399], [4049.09082588, 3382.29455772, 3954.15834981], [4422.29348522, 3549.2372973 , 4066.31337586]], [[4212.20435522, 3476.71136441, 4104.05998443], [4139.66832504, 3371.93545224, 4198.27086105], [4369.19834033, 3797.3129873 , 3828.50248262], ..., [4320.92217993, 3755.9095065 , 4165.54070368], [4231.08308098, 3713.65002654, 3964.55758405], [4274.41509776, 3407.86141075, 3931.29076179]], [[4314.0372303 , 3573.73152574, 4058.10772914], [4114.59770596, 3463.36390945, 4018.74209014], [4464.57525582, 3705.14913054, 4089.96639073], ..., [4504.44403668, 3638.42652115, 3911.16769539], [4256.83526394, 3886.43065114, 3954.52402571], [4256.83526394, 3886.43065114, 3954.52402571]]])
-
helper(chain, draw, replay)float640.7894 0.5491 ... 1.849 -0.6287
array([[[ 0.78936952, 0.5490887 , 0.16650833, ..., 0.08329766, -0.65050062, 0.29929534], [ 1.58545972, -0.65388647, -0.65379105, ..., 1.82456396, -0.19362754, 1.51704283], [-0.95306398, 1.74037107, -1.06139752, ..., 0.49580659, -1.65596289, 0.84096651], ..., [ 0.83862077, 0.72070695, 0.48932378, ..., -0.2039098 , -1.29151567, 1.33576567], [ 0.36248347, 1.46751904, 2.691205 , ..., 1.45830882, -0.35768443, 2.55530514], [ 1.08745544, 1.27714706, 0.61058664, ..., 1.29672473, -0.50342047, 2.35664708]], [[ 1.03478446, 0.34214246, -0.02786286, ..., -1.94344799, -1.74622235, 0.23291672], [-1.13226371, 0.59456024, -0.63842647, ..., 0.36564697, -0.64213818, 1.34387537], [-1.13076226, 0.23707974, -1.6545314 , ..., -0.40823523, -2.01952747, 1.78589938], ..., [ 1.07091423, -0.57066178, 0.11606562, ..., -1.94859916, -0.54562754, 0.47149236], [ 1.73615185, -0.27786063, -0.27283103, ..., -1.55429249, -0.73246607, 0.73910541], [-0.65499351, 1.15010742, 0.65100113, ..., -0.51231793, -0.53271042, 1.44304005]], [[ 0.49833716, -0.51620112, 0.13741193, ..., 0.501271 , 1.09108997, -0.73316122], [-0.25632797, 0.70929871, -0.30192588, ..., 0.74061067, 0.88777097, 0.1928401 ], [ 0.24746999, -0.78483546, 0.05723709, ..., -0.06153329, -2.12186644, 0.20316985], ..., [ 2.14989454, 1.12821892, 0.49014978, ..., 0.70910716, -0.11305358, -0.26844358], [ 0.79712821, -0.98653506, 0.27999657, ..., 0.02147621, 0.45285935, -0.223395 ], [ 1.17281098, -0.88489812, 2.22847395, ..., 0.52398786, -0.19246968, 0.97882991]], [[ 0.15828557, -0.38448579, -0.2247607 , ..., 0.09849247, -0.17279176, 0.79657494], [ 0.3954921 , -0.67537986, -0.07942884, ..., 1.00072493, -0.15591947, 0.71581227], [ 2.66366278, 0.60968907, -0.42128861, ..., -0.09888939, 0.52831846, 0.29878227], ..., [-0.54199678, -0.4028017 , -1.59414456, ..., 2.31936087, 0.07555759, 0.57843756], [ 1.20184084, 0.51496559, 0.73758509, ..., 1.82835713, 1.8487901 , -0.62868765], [ 1.20184084, 0.51496559, 0.73758509, ..., 1.82835713, 1.8487901 , -0.62868765]]])
-
σ(chain, draw, race)float64140.4 252.7 183.9 ... 72.88 55.78
array([[[140.36011672, 252.69076651, 183.8904596 ], [161.51391431, 82.67823509, 212.55352477], [108.91799094, 66.69192358, 88.42575348], ..., [ 31.69679346, 101.09097511, 32.56956811], [148.25558444, 50.48937963, 103.02389557], [170.53973503, 42.45131739, 7.22075687]], [[ 81.39860974, 97.00223551, 5.23367679], [ 28.60567987, 120.58701662, 119.74595091], [ 24.85236716, 47.00875165, 62.47517088], ..., [ 27.08894035, 109.52428738, 89.74107174], [101.36114868, 44.00684249, 63.21585206], [ 45.44140301, 42.24565628, 32.21911083]], [[ 85.15403518, 72.24973621, 101.9819129 ], [ 69.33515359, 91.23129494, 140.7847575 ], [ 45.56776301, 101.34687473, 65.85446419], ..., [ 39.67975301, 28.28842643, 52.09586543], [227.54401728, 35.04100696, 46.01622763], [141.84401665, 229.02600795, 5.01932257]], [[170.73574582, 6.64000168, 256.76699449], [ 77.7241646 , 58.94863209, 63.38158179], [ 92.43500404, 57.06879911, 99.61428138], ..., [ 34.21264796, 52.28070628, 126.57472452], [179.13659055, 72.88096661, 55.7761304 ], [179.13659055, 72.88096661, 55.7761304 ]]])
-
MMR(chain, draw, replay)float643.574e+03 3.514e+03 ... 3.919e+03
array([[[3574.47650843, 3513.75976477, 4079.48244533, ..., 4314.48720853, 4211.49119573, 4103.90070894], [3426.82524018, 3241.68004931, 3815.08489516, ..., 4618.55704024, 4292.59103243, 4276.50328733], [3384.72727505, 3564.35763932, 4141.85534813, ..., 4293.67624891, 4059.3098399 , 4310.07332137], ..., [3677.74379789, 3665.82377497, 3923.71921341, ..., 4244.49172209, 4210.01810344, 3951.28746012], [3675.90453027, 3731.69709048, 4318.79087106, ..., 4537.56294544, 4268.33180356, 4304.78993801], [3726.36593656, 3734.41859557, 4045.85794741, ..., 4599.72656921, 4292.73028387, 4058.46582531]], [[3486.32112569, 3419.1333035 , 4135.21741888, ..., 3982.86934633, 3998.92323941, 4136.58225494], [3533.87103837, 3742.10358598, 3969.14018261, ..., 4149.26695054, 4120.43857122, 4206.51280099], [3623.24589541, 3687.54644032, 3827.74889736, ..., 4184.58134634, 4144.53692 , 4042.69039863], ..., [3547.14865165, 3367.35620982, 4014.82590683, ..., 4013.31809649, 4051.32311109, 4046.72228372], [3458.69711886, 3370.06678874, 3936.91110358, ..., 3891.5459539 , 3974.84722316, 4000.8815279 ], [3521.5666664 , 3597.8243401 , 4087.28805338, ..., 4399.01303974, 4398.08637637, 4112.80684303]], [[3512.71609283, 3439.41596971, 4118.0735164 , ..., 4254.88960386, 4305.11506861, 4029.29080031], [3348.55031945, 3436.64569193, 4155.76429966, ..., 4191.01867944, 4201.22206179, 4225.4198078 ], [3822.39329786, 3717.77236616, 3832.27180076, ..., 4366.39440592, 4272.50963317, 3841.88212429], ..., [3816.72664015, 3787.82504444, 4191.07548085, ..., 4349.0593768 , 4316.43624179, 4151.55590297], [3741.58220182, 3679.08084456, 3977.44197015, ..., 4235.96986521, 4334.12851785, 3954.27778896], [3676.46562714, 3205.19672793, 3942.47619137, ..., 4348.73963984, 4247.11442498, 3936.20382483]], [[3574.78254219, 3571.17853943, 4000.39660036, ..., 4330.853415 , 4284.53550087, 4262.64188132], [3486.67762767, 3423.55119059, 4013.70776491, ..., 4192.37821489, 4102.47899548, 4064.11140391], [3857.16116685, 3739.94335347, 4048.00002814, ..., 4455.43441496, 4513.41037517, 4119.72937191], ..., [3610.09054683, 3617.36776396, 3709.38928727, ..., 4583.79551357, 4507.02906206, 3984.38327026], [3974.02197305, 3923.96184115, 3995.66366771, ..., 4584.36092651, 4588.0212183 , 3919.4582613 ], [3974.02197305, 3923.96184115, 3995.66366771, ..., 4584.36092651, 4588.0212183 , 3919.4582613 ]]])
-
MMR_diff(chain, draw, replay)float64-173.5 -432.2 ... 651.0 -320.5
array([[[-173.52349157, -432.24023523, 277.48244533, ..., 337.48720853, 274.49119573, -136.09929106], [-321.17475982, -704.31995069, 13.08489516, ..., 641.55704024, 355.59103243, 36.50328733], [-363.27272495, -381.64236068, 339.85534813, ..., 316.67624891, 122.3098399 , 70.07332137], ..., [ -70.25620211, -280.17622503, 121.71921341, ..., 267.49172209, 273.01810344, -288.71253988], [ -72.09546973, -214.30290952, 516.79087106, ..., 560.56294544, 331.33180356, 64.78993801], [ -21.63406344, -211.58140443, 243.85794741, ..., 622.72656921, 355.73028387, -181.53417469]], [[-261.67887431, -526.8666965 , 333.21741888, ..., 5.86934633, 61.92323941, -103.41774506], [-214.12896163, -203.89641402, 167.14018261, ..., 172.26695054, 183.43857122, -33.48719901], [-124.75410459, -258.45355968, 25.74889736, ..., 207.58134634, 207.53692 , -197.30960137], ..., [-200.85134835, -578.64379018, 212.82590683, ..., 36.31809649, 114.32311109, -193.27771628], [-289.30288114, -575.93321126, 134.91110358, ..., -85.4540461 , 37.84722316, -239.1184721 ], [-226.4333336 , -348.1756599 , 285.28805338, ..., 422.01303974, 461.08637637, -127.19315697]], [[-235.28390717, -506.58403029, 316.0735164 , ..., 277.88960386, 368.11506861, -210.70919969], [-399.44968055, -509.35430807, 353.76429966, ..., 214.01867944, 264.22206179, -14.5801922 ], [ 74.39329786, -228.22763384, 30.27180076, ..., 389.39440592, 335.50963317, -398.11787571], ..., [ 68.72664015, -158.17495556, 389.07548085, ..., 372.0593768 , 379.43624179, -88.44409703], [ -6.41779818, -266.91915544, 175.44197015, ..., 258.96986521, 397.12851785, -285.72221104], [ -71.53437286, -740.80327207, 140.47619137, ..., 371.73963984, 310.11442498, -303.79617517]], [[-173.21745781, -374.82146057, 198.39660036, ..., 353.853415 , 347.53550087, 22.64188132], [-261.32237233, -522.44880941, 211.70776491, ..., 215.37821489, 165.47899548, -175.88859609], [ 109.16116685, -206.05664653, 246.00002814, ..., 478.43441496, 576.41037517, -120.27062809], ..., [-137.90945317, -328.63223604, -92.61071273, ..., 606.79551357, 570.02906206, -255.61672974], [ 226.02197305, -22.03815885, 193.66366771, ..., 607.36092651, 651.0212183 , -320.5417387 ], [ 226.02197305, -22.03815885, 193.66366771, ..., 607.36092651, 651.0212183 , -320.5417387 ]]])
-
winrate(chain, draw, replay)float640.3884 0.244 ... 0.846 0.3018
array([[[0.38840135, 0.24397871, 0.67393861, ..., 0.70745598, 0.67221636, 0.41190061], [0.30145528, 0.13670806, 0.50855856, ..., 0.84272985, 0.71716256, 0.52386025], [0.27877826, 0.26921766, 0.70873675, ..., 0.69606027, 0.57933223, 0.54571002], ..., [0.45417135, 0.32451442, 0.57895556, ..., 0.66816823, 0.67136651, 0.31963761], [0.45297858, 0.36337639, 0.79449259, ..., 0.81256357, 0.70411153, 0.5422807 ], [0.485852 , 0.36502532, 0.65431758, ..., 0.83608882, 0.71723646, 0.38343406]], [[0.33521265, 0.20123624, 0.7051384 , ..., 0.50383932, 0.5404183 , 0.43275976], [0.36348169, 0.3696986 , 0.60762375, ..., 0.61081735, 0.61774329, 0.47810857], [0.41910992, 0.3370959 , 0.51683711, ..., 0.63254534, 0.63251832, 0.37372364], ..., [0.37155715, 0.18033725, 0.6357291 , ..., 0.52373938, 0.57423102, 0.37619612], [0.31930178, 0.181388 , 0.58734606, ..., 0.44433253, 0.5247373 , 0.34849273], [0.35606613, 0.28678965, 0.67841064, ..., 0.75105153, 0.76967316, 0.41755699]], [[0.35077423, 0.20990227, 0.69572652, ..., 0.67417267, 0.7237621 , 0.36555445], [0.26014989, 0.20870266, 0.71619203, ..., 0.63645155, 0.66626863, 0.49046362], [0.54851081, 0.35499039, 0.51979175, ..., 0.73475424, 0.70638391, 0.26082117], ..., [0.54483632, 0.39798298, 0.73459157, ..., 0.72582073, 0.72964516, 0.4424017 ], [0.49580194, 0.33216402, 0.61279042, ..., 0.66320592, 0.73867943, 0.32134157], [0.45334239, 0.12582742, 0.59087079, ..., 0.72565421, 0.69241571, 0.31111641]], [[0.38859158, 0.27274337, 0.62694182, ..., 0.71623942, 0.71286761, 0.51480671], [0.33542055, 0.20310077, 0.63505131, ..., 0.63727424, 0.60658696, 0.38693233], [0.57092553, 0.36838244, 0.65558424, ..., 0.7776215 , 0.81879732, 0.42196867], ..., [0.41075375, 0.29736221, 0.43971402, ..., 0.83029583, 0.81630679, 0.33875661], [0.64368704, 0.48558788, 0.62404084, ..., 0.83050419, 0.84598416, 0.30180419], [0.64368704, 0.48558788, 0.62404084, ..., 0.83050419, 0.84598416, 0.30180419]]])
-
-
- created_at :
- 2020-06-20T10:35:36.555638
- arviz_version :
- 0.8.3
- inference_library :
- pymc3
- inference_library_version :
- 3.9.1
- sampling_time :
- 18.211029291152954
- tuning_steps :
- 2000
xarray.Dataset
-
-
-
- chain: 4
- draw: 2000
- race: 3
- replay: 138
-
-
chain(chain)int640 1 2 3
array([0, 1, 2, 3])
-
draw(draw)int640 1 2 3 4 ... 1996 1997 1998 1999
array([ 0, 1, 2, ..., 1997, 1998, 1999])
-
replay(replay)int648 325 54 346 138 ... 25 208 59 364
array([ 8, 325, 54, 346, 138, 405, 20, 129, 104, 46, 302, 408, 101, 219, 316, 126, 98, 231, 241, 385, 193, 198, 90, 329, 137, 200, 80, 355, 317, 33, 64, 213, 368, 49, 435, 134, 254, 330, 60, 39, 218, 109, 301, 133, 328, 181, 156, 395, 43, 249, 27, 153, 211, 420, 366, 186, 163, 63, 202, 45, 69, 31, 167, 177, 95, 151, 392, 387, 18, 286, 102, 290, 195, 428, 403, 97, 406, 412, 374, 263, 371, 41, 212, 52, 238, 345, 4, 117, 407, 56, 103, 118, 319, 57, 128, 294, 15, 273, 327, 281, 378, 121, 113, 284, 422, 389, 216, 58, 418, 309, 123, 116, 222, 122, 11, 361, 179, 255, 239, 225, 381, 424, 343, 287, 341, 184, 380, 196, 174, 306, 252, 148, 416, 391, 25, 208, 59, 364])
-
race(race)<U7'Terran' 'Protoss' 'Zerg'
array(['Terran', 'Protoss', 'Zerg'], dtype='<U7')
-
-
-
win(chain, draw, replay)int640 0 1 1 0 0 1 0 ... 1 0 1 0 1 1 1 0
array([[[0, 0, 1, ..., 1, 0, 0], [1, 0, 0, ..., 1, 0, 0], [0, 0, 1, ..., 1, 0, 0], ..., [1, 0, 0, ..., 1, 1, 1], [1, 1, 1, ..., 1, 1, 1], [0, 0, 1, ..., 1, 1, 1]], [[1, 0, 0, ..., 0, 1, 0], [1, 0, 1, ..., 1, 1, 0], [0, 0, 1, ..., 1, 1, 0], ..., [1, 0, 1, ..., 0, 0, 1], [1, 0, 1, ..., 1, 1, 0], [1, 0, 1, ..., 0, 1, 0]], [[1, 1, 0, ..., 0, 0, 0], [1, 0, 1, ..., 1, 1, 0], [0, 1, 0, ..., 1, 1, 1], ..., [1, 0, 0, ..., 1, 1, 0], [0, 1, 1, ..., 1, 1, 0], [1, 0, 0, ..., 1, 1, 0]], [[0, 0, 1, ..., 1, 1, 1], [0, 0, 1, ..., 1, 0, 0], [0, 1, 0, ..., 1, 0, 1], ..., [0, 0, 0, ..., 0, 1, 1], [1, 0, 0, ..., 1, 0, 0], [0, 0, 0, ..., 1, 1, 0]]])
-
μ(chain, draw, race)float644.303e+03 3.375e+03 ... 3.955e+03
array([[[4302.79553938, 3375.01011989, 4048.86315149], [4323.86457361, 3295.74222841, 3954.05048638], [4239.6739907 , 3448.28894514, 4235.71022401], ..., [4250.95500899, 3592.96680679, 3907.78214913], [4321.36051851, 3657.60296464, 4041.53244852], [4378.58347689, 3680.20202035, 4041.44904971]], [[4141.06331072, 3385.94471985, 4135.36324411], [4138.80737046, 3670.4073407 , 4045.58916728], [4194.72695819, 3676.40161761, 3931.11602939], ..., [4066.10358294, 3429.85753425, 4004.41005399], [4049.09082588, 3382.29455772, 3954.15834981], [4422.29348522, 3549.2372973 , 4066.31337586]], [[4212.20435522, 3476.71136441, 4104.05998443], [4139.66832504, 3371.93545224, 4198.27086105], [4369.19834033, 3797.3129873 , 3828.50248262], ..., [4320.92217993, 3755.9095065 , 4165.54070368], [4231.08308098, 3713.65002654, 3964.55758405], [4274.41509776, 3407.86141075, 3931.29076179]], [[4314.0372303 , 3573.73152574, 4058.10772914], [4114.59770596, 3463.36390945, 4018.74209014], [4464.57525582, 3705.14913054, 4089.96639073], ..., [4504.44403668, 3638.42652115, 3911.16769539], [4256.83526394, 3886.43065114, 3954.52402571], [4256.83526394, 3886.43065114, 3954.52402571]]])
-
σ(chain, draw, race)float64140.4 252.7 183.9 ... 72.88 55.78
array([[[140.36011672, 252.69076651, 183.8904596 ], [161.51391431, 82.67823509, 212.55352477], [108.91799094, 66.69192358, 88.42575348], ..., [ 31.69679346, 101.09097511, 32.56956811], [148.25558444, 50.48937963, 103.02389557], [170.53973503, 42.45131739, 7.22075687]], [[ 81.39860974, 97.00223551, 5.23367679], [ 28.60567987, 120.58701662, 119.74595091], [ 24.85236716, 47.00875165, 62.47517088], ..., [ 27.08894035, 109.52428738, 89.74107174], [101.36114868, 44.00684249, 63.21585206], [ 45.44140301, 42.24565628, 32.21911083]], [[ 85.15403518, 72.24973621, 101.9819129 ], [ 69.33515359, 91.23129494, 140.7847575 ], [ 45.56776301, 101.34687473, 65.85446419], ..., [ 39.67975301, 28.28842643, 52.09586543], [227.54401728, 35.04100696, 46.01622763], [141.84401665, 229.02600795, 5.01932257]], [[170.73574582, 6.64000168, 256.76699449], [ 77.7241646 , 58.94863209, 63.38158179], [ 92.43500404, 57.06879911, 99.61428138], ..., [ 34.21264796, 52.28070628, 126.57472452], [179.13659055, 72.88096661, 55.7761304 ], [179.13659055, 72.88096661, 55.7761304 ]]])
-
winrate(chain, draw, replay)float640.3884 0.244 ... 0.846 0.3018
array([[[0.38840135, 0.24397871, 0.67393861, ..., 0.70745598, 0.67221636, 0.41190061], [0.30145528, 0.13670806, 0.50855856, ..., 0.84272985, 0.71716256, 0.52386025], [0.27877826, 0.26921766, 0.70873675, ..., 0.69606027, 0.57933223, 0.54571002], ..., [0.45417135, 0.32451442, 0.57895556, ..., 0.66816823, 0.67136651, 0.31963761], [0.45297858, 0.36337639, 0.79449259, ..., 0.81256357, 0.70411153, 0.5422807 ], [0.485852 , 0.36502532, 0.65431758, ..., 0.83608882, 0.71723646, 0.38343406]], [[0.33521265, 0.20123624, 0.7051384 , ..., 0.50383932, 0.5404183 , 0.43275976], [0.36348169, 0.3696986 , 0.60762375, ..., 0.61081735, 0.61774329, 0.47810857], [0.41910992, 0.3370959 , 0.51683711, ..., 0.63254534, 0.63251832, 0.37372364], ..., [0.37155715, 0.18033725, 0.6357291 , ..., 0.52373938, 0.57423102, 0.37619612], [0.31930178, 0.181388 , 0.58734606, ..., 0.44433253, 0.5247373 , 0.34849273], [0.35606613, 0.28678965, 0.67841064, ..., 0.75105153, 0.76967316, 0.41755699]], [[0.35077423, 0.20990227, 0.69572652, ..., 0.67417267, 0.7237621 , 0.36555445], [0.26014989, 0.20870266, 0.71619203, ..., 0.63645155, 0.66626863, 0.49046362], [0.54851081, 0.35499039, 0.51979175, ..., 0.73475424, 0.70638391, 0.26082117], ..., [0.54483632, 0.39798298, 0.73459157, ..., 0.72582073, 0.72964516, 0.4424017 ], [0.49580194, 0.33216402, 0.61279042, ..., 0.66320592, 0.73867943, 0.32134157], [0.45334239, 0.12582742, 0.59087079, ..., 0.72565421, 0.69241571, 0.31111641]], [[0.38859158, 0.27274337, 0.62694182, ..., 0.71623942, 0.71286761, 0.51480671], [0.33542055, 0.20310077, 0.63505131, ..., 0.63727424, 0.60658696, 0.38693233], [0.57092553, 0.36838244, 0.65558424, ..., 0.7776215 , 0.81879732, 0.42196867], ..., [0.41075375, 0.29736221, 0.43971402, ..., 0.83029583, 0.81630679, 0.33875661], [0.64368704, 0.48558788, 0.62404084, ..., 0.83050419, 0.84598416, 0.30180419], [0.64368704, 0.48558788, 0.62404084, ..., 0.83050419, 0.84598416, 0.30180419]]])
-
-
- created_at :
- 2020-06-20T10:35:37.090348
- arviz_version :
- 0.8.3
- inference_library :
- pymc3
- inference_library_version :
- 3.9.1
xarray.Dataset
-
-
-
- chain: 4
- draw: 2000
- replay: 138
-
-
chain(chain)int640 1 2 3
array([0, 1, 2, 3])
-
draw(draw)int640 1 2 3 4 ... 1996 1997 1998 1999
array([ 0, 1, 2, ..., 1997, 1998, 1999])
-
replay(replay)int648 325 54 346 138 ... 25 208 59 364
array([ 8, 325, 54, 346, 138, 405, 20, 129, 104, 46, 302, 408, 101, 219, 316, 126, 98, 231, 241, 385, 193, 198, 90, 329, 137, 200, 80, 355, 317, 33, 64, 213, 368, 49, 435, 134, 254, 330, 60, 39, 218, 109, 301, 133, 328, 181, 156, 395, 43, 249, 27, 153, 211, 420, 366, 186, 163, 63, 202, 45, 69, 31, 167, 177, 95, 151, 392, 387, 18, 286, 102, 290, 195, 428, 403, 97, 406, 412, 374, 263, 371, 41, 212, 52, 238, 345, 4, 117, 407, 56, 103, 118, 319, 57, 128, 294, 15, 273, 327, 281, 378, 121, 113, 284, 422, 389, 216, 58, 418, 309, 123, 116, 222, 122, 11, 361, 179, 255, 239, 225, 381, 424, 343, 287, 341, 184, 380, 196, 174, 306, 252, 148, 416, 391, 25, 208, 59, 364])
-
-
-
win(chain, draw, replay)float64-0.9457 -0.2797 ... -0.1673 -1.198
array([[[-0.94571606, -0.27968575, -1.1206696 , ..., -0.34607987, -0.39717502, -0.88697319], [-1.1991336 , -0.14700236, -0.7104125 , ..., -0.17110883, -0.33245274, -0.64653033], [-1.27733859, -0.31363962, -1.23352778, ..., -0.36231903, -0.54587916, -0.60566754], ..., [-0.78928073, -0.39232348, -0.86501689, ..., -0.40321529, -0.39844008, -1.14056741], [-0.79191044, -0.45157668, -1.58227317, ..., -0.20756113, -0.35081851, -0.61197151], [-0.72185124, -0.45417016, -1.0622348 , ..., -0.17902043, -0.3323497 , -0.95858761]], [[-1.09299019, -0.22469005, -1.2212492 , ..., -0.68549787, -0.61541181, -0.83757254], [-1.01202636, -0.46155716, -0.93553407, ..., -0.4929573 , -0.48168229, -0.73791744], [-0.86962207, -0.41112495, -0.72740143, ..., -0.45800338, -0.4580461 , -0.9842387 ], ..., [-0.99005259, -0.1988623 , -1.00985747, ..., -0.64676108, -0.5547235 , -0.97764469], [-1.1416186 , -0.20014505, -0.88514597, ..., -0.81118206, -0.64485752, -1.0541379 ], [-1.03263881, -0.33797888, -1.13447981, ..., -0.28628102, -0.26178932, -0.87333424]], [[-1.04761249, -0.23559864, -1.18982837, ..., -0.39426901, -0.32329253, -1.00634002], [-1.34649731, -0.23408148, -1.25945741, ..., -0.45184699, -0.40606233, -0.71240417], [-0.60054829, -0.43849007, -0.73353543, ..., -0.30821921, -0.34759641, -1.34392027], ..., [-0.60726986, -0.50746956, -1.3264854 , ..., -0.32045223, -0.31519695, -0.81553699], [-0.70157875, -0.40371268, -0.94878918, ..., -0.41066975, -0.30289124, -1.13525063], [-0.7911076 , -0.13447746, -0.89372426, ..., -0.32068168, -0.36756877, -1.16758814]], [[-0.9452264 , -0.31847587, -0.98602091, ..., -0.33374078, -0.33845955, -0.66396377], [-1.09237016, -0.22702705, -1.00799851, ..., -0.45055519, -0.49990719, -0.94950547], [-0.56049651, -0.45947119, -1.06590574, ..., -0.25151537, -0.1999187 , -0.86282421], ..., [-0.88976139, -0.35291376, -0.57930794, ..., -0.18597321, -0.20296503, -1.08247338], [-0.44054263, -0.66473054, -0.97827477, ..., -0.1857223 , -0.16725464, -1.19797685], [-0.44054263, -0.66473054, -0.97827477, ..., -0.1857223 , -0.16725464, -1.19797685]]])
-
-
- created_at :
- 2020-06-20T10:35:37.087198
- arviz_version :
- 0.8.3
- inference_library :
- pymc3
- inference_library_version :
- 3.9.1
xarray.Dataset
-
-
-
- chain: 4
- draw: 2000
-
-
chain(chain)int640 1 2 3
array([0, 1, 2, 3])
-
draw(draw)int640 1 2 3 4 ... 1996 1997 1998 1999
array([ 0, 1, 2, ..., 1997, 1998, 1999])
-
-
-
step_size_bar(chain, draw)float640.4374 0.4374 ... 0.4903 0.4903
array([[0.4374115 , 0.4374115 , 0.4374115 , ..., 0.4374115 , 0.4374115 , 0.4374115 ], [0.45402308, 0.45402308, 0.45402308, ..., 0.45402308, 0.45402308, 0.45402308], [0.41388167, 0.41388167, 0.41388167, ..., 0.41388167, 0.41388167, 0.41388167], [0.49034327, 0.49034327, 0.49034327, ..., 0.49034327, 0.49034327, 0.49034327]])
-
energy_error(chain, draw)float64-1.016 0.3132 ... 0.6163 0.0
array([[-1.01642236, 0.31318099, -0.14019916, ..., -0.00712864, -0.41741377, -0.08385882], [-0.38042493, -1.50079731, -0.16318778, ..., -0.24735973, -0.2886005 , -0.17678952], [-2.01224887, -0.89066887, 0.47713108, ..., -1.70039887, 0.31851951, 0.11482341], [ 1.43938357, 0.2229521 , 0.32288281, ..., -0.04535445, 0.61629039, 0. ]])
-
step_size(chain, draw)float640.4386 0.4386 ... 0.425 0.425
array([[0.43855251, 0.43855251, 0.43855251, ..., 0.43855251, 0.43855251, 0.43855251], [0.48249608, 0.48249608, 0.48249608, ..., 0.48249608, 0.48249608, 0.48249608], [0.44391835, 0.44391835, 0.44391835, ..., 0.44391835, 0.44391835, 0.44391835], [0.42501071, 0.42501071, 0.42501071, ..., 0.42501071, 0.42501071, 0.42501071]])
-
max_energy_error(chain, draw)float64-1.287 1.11 -0.4291 ... 1.535 5.196
array([[-1.28737009, 1.11007453, -0.42910467, ..., 0.91347544, 0.52800764, 0.47859559], [-0.48591337, -1.50079731, 2.16250099, ..., -0.24735973, 3.77015272, -0.76572754], [-2.328056 , 1.01810228, 0.80442549, ..., -1.70039887, -0.55804924, 74.25084326], [38.71084628, 1.20965313, 0.49249056, ..., 1.35746487, 1.53478843, 5.19565611]])
-
diverging(chain, draw)boolFalse False False ... False False
array([[False, False, False, ..., False, False, False], [False, False, False, ..., False, False, False], [False, False, False, ..., False, False, False], [False, False, False, ..., False, False, False]])
-
depth(chain, draw)int644 3 3 4 4 3 3 3 ... 3 3 3 3 3 3 3 3
array([[4, 3, 3, ..., 3, 3, 4], [3, 3, 3, ..., 3, 3, 3], [4, 3, 3, ..., 4, 4, 4], [3, 3, 3, ..., 3, 3, 3]])
-
mean_tree_accept(chain, draw)float640.9054 0.7214 ... 0.4055 0.07633
array([[0.90541692, 0.72142566, 0.93134252, ..., 0.85369685, 0.9554831 , 0.86302654], [0.99759711, 0.99963932, 0.88165678, ..., 0.96310244, 0.78523323, 0.97673572], [0.99816125, 0.88683585, 0.75217549, ..., 0.96863687, 0.92512264, 0.86677524], [0.40260816, 0.83998264, 0.84497975, ..., 0.76832421, 0.4055396 , 0.07633182]])
-
energy(chain, draw)float64356.5 384.7 373.6 ... 368.5 373.4
array([[356.54520829, 384.72052031, 373.61948961, ..., 390.95997962, 395.89754948, 395.0117275 ], [410.94614041, 383.5146222 , 361.83684381, ..., 373.49242948, 364.71453224, 360.19568382], [391.28602749, 376.75745726, 387.21260602, ..., 390.37738103, 377.07691302, 389.32502442], [358.39179418, 363.878841 , 377.07222589, ..., 378.63812646, 368.46859802, 373.42582631]])
-
lp(chain, draw)float64-302.1 -305.8 ... -302.2 -302.2
array([[-302.08558028, -305.77270943, -300.43477915, ..., -320.70266421, -315.38015305, -311.86956419], [-324.9312429 , -298.99886287, -298.64220017, ..., -301.73942136, -301.70998423, -295.83681186], [-310.88193965, -301.57205804, -313.61330102, ..., -309.33294253, -310.28663405, -315.73379124], [-294.91496388, -306.34122894, -309.38563566, ..., -304.01227822, -302.20722956, -302.20722956]])
-
tree_size(chain, draw)float6415.0 7.0 7.0 15.0 ... 7.0 7.0 7.0
array([[15., 7., 7., ..., 7., 7., 15.], [ 7., 7., 7., ..., 7., 7., 7.], [15., 7., 7., ..., 15., 15., 15.], [ 7., 7., 7., ..., 7., 7., 7.]])
-
-
- created_at :
- 2020-06-20T10:35:36.562793
- arviz_version :
- 0.8.3
- inference_library :
- pymc3
- inference_library_version :
- 3.9.1
- sampling_time :
- 18.211029291152954
- tuning_steps :
- 2000
xarray.Dataset
-
-
-
- chain: 1
- draw: 2000
- race: 3
- replay: 138
-
-
chain(chain)int640
array([0])
-
draw(draw)int640 1 2 3 4 ... 1996 1997 1998 1999
array([ 0, 1, 2, ..., 1997, 1998, 1999])
-
race(race)<U7'Terran' 'Protoss' 'Zerg'
array(['Terran', 'Protoss', 'Zerg'], dtype='<U7')
-
replay(replay)int648 325 54 346 138 ... 25 208 59 364
array([ 8, 325, 54, 346, 138, 405, 20, 129, 104, 46, 302, 408, 101, 219, 316, 126, 98, 231, 241, 385, 193, 198, 90, 329, 137, 200, 80, 355, 317, 33, 64, 213, 368, 49, 435, 134, 254, 330, 60, 39, 218, 109, 301, 133, 328, 181, 156, 395, 43, 249, 27, 153, 211, 420, 366, 186, 163, 63, 202, 45, 69, 31, 167, 177, 95, 151, 392, 387, 18, 286, 102, 290, 195, 428, 403, 97, 406, 412, 374, 263, 371, 41, 212, 52, 238, 345, 4, 117, 407, 56, 103, 118, 319, 57, 128, 294, 15, 273, 327, 281, 378, 121, 113, 284, 422, 389, 216, 58, 418, 309, 123, 116, 222, 122, 11, 361, 179, 255, 239, 225, 381, 424, 343, 287, 341, 184, 380, 196, 174, 306, 252, 148, 416, 391, 25, 208, 59, 364])
-
-
-
μ(chain, draw, race)float644.487e+03 3.816e+03 ... 4.079e+03
array([[[4487.3036091 , 3816.4730759 , 3841.54847432], [3678.10941335, 4259.6222888 , 3309.53839094], [4523.44352926, 3771.63792973, 4095.71172882], ..., [3793.7352304 , 3936.80991215, 3897.05804332], [4287.27113474, 4067.38137671, 3856.85673442], [3792.63645476, 3763.38063103, 4079.25539045]]])
-
σ(chain, draw, race)float6410.45 53.54 169.3 ... 210.5 146.2
array([[[ 10.45030045, 53.53690568, 169.32540234], [ 23.57617238, 10.93212404, 83.18141646], [125.6331917 , 94.68059879, 168.58509729], ..., [ 54.01739286, 112.35908554, 107.02562404], [144.26356103, 77.82838645, 9.93539497], [ 81.81960096, 210.54208204, 146.24259794]]])
-
winrate(chain, draw, replay)float640.4452 0.4443 ... 0.4049 0.3675
array([[[0.44516586, 0.44432904, 0.49543905, ..., 0.79673329, 0.80140205, 0.28212387], [0.7972168 , 0.68737573, 0.21071926, ..., 0.31760374, 0.3228919 , 0.08274033], [0.46947403, 0.43159697, 0.78947532, ..., 0.82439068, 0.80940984, 0.32718542], ..., [0.74439685, 0.57396175, 0.48594935, ..., 0.37104134, 0.37754386, 0.3333573 ], [0.70701564, 0.56655281, 0.54098671, ..., 0.57389202, 0.76269576, 0.26823939], [0.59530994, 0.28697282, 0.72372083, ..., 0.36763267, 0.40492739, 0.36749525]]])
-
-
- created_at :
- 2020-06-20T10:35:37.093692
- arviz_version :
- 0.8.3
- inference_library :
- pymc3
- inference_library_version :
- 3.9.1
xarray.Dataset
-
-
-
- chain: 1
- draw: 2000
- replay: 138
-
-
chain(chain)int640
array([0])
-
draw(draw)int640 1 2 3 4 ... 1996 1997 1998 1999
array([ 0, 1, 2, ..., 1997, 1998, 1999])
-
replay(replay)int648 325 54 346 138 ... 25 208 59 364
array([ 8, 325, 54, 346, 138, 405, 20, 129, 104, 46, 302, 408, 101, 219, 316, 126, 98, 231, 241, 385, 193, 198, 90, 329, 137, 200, 80, 355, 317, 33, 64, 213, 368, 49, 435, 134, 254, 330, 60, 39, 218, 109, 301, 133, 328, 181, 156, 395, 43, 249, 27, 153, 211, 420, 366, 186, 163, 63, 202, 45, 69, 31, 167, 177, 95, 151, 392, 387, 18, 286, 102, 290, 195, 428, 403, 97, 406, 412, 374, 263, 371, 41, 212, 52, 238, 345, 4, 117, 407, 56, 103, 118, 319, 57, 128, 294, 15, 273, 327, 281, 378, 121, 113, 284, 422, 389, 216, 58, 418, 309, 123, 116, 222, 122, 11, 361, 179, 255, 239, 225, 381, 424, 343, 287, 341, 184, 380, 196, 174, 306, 252, 148, 416, 391, 25, 208, 59, 364])
-
-
-
win(chain, draw, replay)int641 1 0 0 1 1 1 0 ... 0 0 0 1 1 0 0 0
array([[[1, 1, 0, ..., 1, 1, 0], [1, 1, 0, ..., 0, 0, 0], [1, 0, 0, ..., 1, 1, 0], ..., [1, 1, 0, ..., 0, 0, 0], [1, 1, 0, ..., 1, 0, 0], [0, 0, 0, ..., 0, 0, 0]]])
-
-
- created_at :
- 2020-06-20T10:35:37.095318
- arviz_version :
- 0.8.3
- inference_library :
- pymc3
- inference_library_version :
- 3.9.1
xarray.Dataset
-
-
-
- replay: 138
-
-
replay(replay)int648 325 54 346 138 ... 25 208 59 364
array([ 8, 325, 54, 346, 138, 405, 20, 129, 104, 46, 302, 408, 101, 219, 316, 126, 98, 231, 241, 385, 193, 198, 90, 329, 137, 200, 80, 355, 317, 33, 64, 213, 368, 49, 435, 134, 254, 330, 60, 39, 218, 109, 301, 133, 328, 181, 156, 395, 43, 249, 27, 153, 211, 420, 366, 186, 163, 63, 202, 45, 69, 31, 167, 177, 95, 151, 392, 387, 18, 286, 102, 290, 195, 428, 403, 97, 406, 412, 374, 263, 371, 41, 212, 52, 238, 345, 4, 117, 407, 56, 103, 118, 319, 57, 128, 294, 15, 273, 327, 281, 378, 121, 113, 284, 422, 389, 216, 58, 418, 309, 123, 116, 222, 122, 11, 361, 179, 255, 239, 225, 381, 424, 343, 287, 341, 184, 380, 196, 174, 306, 252, 148, 416, 391, 25, 208, 59, 364])
-
-
-
win(replay)float641.0 0.0 0.0 1.0 ... 1.0 1.0 1.0 1.0
array([1., 0., 0., 1., 1., 0., 1., 0., 1., 0., 0., 1., 0., 0., 0., 0., 1., 1., 0., 0., 0., 1., 0., 0., 1., 0., 0., 1., 1., 1., 1., 1., 1., 0., 0., 0., 1., 0., 1., 1., 1., 1., 1., 1., 1., 0., 1., 0., 1., 0., 1., 0., 1., 0., 1., 1., 1., 1., 0., 0., 1., 0., 0., 1., 0., 0., 0., 1., 1., 1., 0., 1., 1., 0., 0., 1., 1., 1., 1., 0., 1., 0., 1., 1., 0., 0., 1., 0., 1., 0., 1., 1., 0., 1., 0., 0., 0., 0., 1., 0., 0., 0., 1., 0., 1., 0., 1., 0., 1., 0., 1., 0., 1., 0., 0., 0., 1., 1., 1., 0., 0., 0., 0., 1., 1., 0., 1., 0., 0., 1., 1., 1., 1., 0., 1., 1., 1., 1.])
-
-
- created_at :
- 2020-06-20T10:35:37.096098
- arviz_version :
- 0.8.3
- inference_library :
- pymc3
- inference_library_version :
- 3.9.1
xarray.Dataset
-
-
-
- replay: 138
-
-
replay(replay)int648 325 54 346 138 ... 25 208 59 364
array([ 8, 325, 54, 346, 138, 405, 20, 129, 104, 46, 302, 408, 101, 219, 316, 126, 98, 231, 241, 385, 193, 198, 90, 329, 137, 200, 80, 355, 317, 33, 64, 213, 368, 49, 435, 134, 254, 330, 60, 39, 218, 109, 301, 133, 328, 181, 156, 395, 43, 249, 27, 153, 211, 420, 366, 186, 163, 63, 202, 45, 69, 31, 167, 177, 95, 151, 392, 387, 18, 286, 102, 290, 195, 428, 403, 97, 406, 412, 374, 263, 371, 41, 212, 52, 238, 345, 4, 117, 407, 56, 103, 118, 319, 57, 128, 294, 15, 273, 327, 281, 378, 121, 113, 284, 422, 389, 216, 58, 418, 309, 123, 116, 222, 122, 11, 361, 179, 255, 239, 225, 381, 424, 343, 287, 341, 184, 380, 196, 174, 306, 252, 148, 416, 391, 25, 208, 59, 364])
-
-
-
enemy_race(replay)int321 1 2 2 1 1 2 1 ... 2 2 0 2 2 0 0 2
array([1, 1, 2, 2, 1, 1, 2, 1, 1, 2, 2, 2, 1, 1, 0, 2, 2, 0, 0, 2, 0, 2, 2, 1, 1, 2, 2, 0, 0, 0, 0, 0, 0, 2, 0, 1, 2, 1, 1, 2, 2, 2, 1, 0, 0, 1, 0, 2, 2, 2, 1, 0, 2, 1, 0, 2, 1, 0, 1, 1, 2, 2, 2, 2, 1, 2, 1, 2, 0, 0, 0, 1, 0, 2, 2, 0, 0, 0, 0, 1, 0, 1, 2, 0, 1, 2, 2, 1, 2, 1, 0, 2, 2, 2, 1, 0, 0, 2, 1, 1, 1, 1, 0, 0, 0, 1, 0, 2, 0, 2, 0, 1, 0, 0, 2, 0, 0, 0, 2, 1, 0, 1, 1, 0, 2, 1, 0, 2, 0, 1, 2, 2, 0, 2, 2, 0, 0, 2], dtype=int32)
-
enemy_mmr(replay)int323748 3946 3802 ... 3977 3937 4240
array([3748, 3946, 3802, 3816, 3701, 3861, 3907, 4066, 3757, 4350, 3964, 3894, 3932, 3952, 3955, 3912, 3744, 3492, 4642, 3756, 3664, 3769, 3871, 3631, 3631, 3685, 3626, 3647, 3626, 3736, 3784, 3856, 3866, 3847, 3883, 3761, 3686, 3827, 3771, 3752, 3803, 3922, 3888, 3900, 3962, 3992, 3798, 3797, 3790, 3966, 3881, 3823, 3904, 3920, 3907, 3961, 3874, 3570, 4359, 4009, 3859, 3988, 3885, 4020, 3935, 3908, 4596, 4851, 3941, 3808, 3942, 3854, 3956, 3976, 3998, 3953, 4068, 3960, 4016, 4055, 3925, 4104, 3989, 4001, 4237, 4161, 4024, 4108, 4125, 4028, 4012, 3992, 4144, 4029, 4051, 4131, 4407, 4037, 3965, 3944, 4169, 3882, 4018, 4066, 3828, 3957, 4181, 4160, 3886, 3963, 3875, 3956, 3976, 3953, 3757, 4007, 3911, 3967, 4935, 3898, 3880, 3927, 3948, 3833, 3894, 3782, 3786, 3845, 3863, 3966, 3864, 3843, 3840, 4033, 3914, 3977, 3937, 4240], dtype=int32)
-
-
- created_at :
- 2020-06-20T10:35:37.097169
- arviz_version :
- 0.8.3
- inference_library :
- pymc3
- inference_library_version :
- 3.9.1
xarray.Dataset
-
That InferenceData
ArviZ object lets us see everything the model picked up. It's really, really neat.
We had a few divergences; not too many, though Let's take a look at how the sampling went:
var_names = "μ σ".split()
az.plot_trace(output, var_names=var_names);
az.summary(output, var_names = var_names)
mean | sd | hdi_3% | hdi_97% | mcse_mean | mcse_sd | ess_mean | ess_sd | ess_bulk | ess_tail | r_hat | |
---|---|---|---|---|---|---|---|---|---|---|---|
μ[0] | 4275.797 | 116.495 | 4066.919 | 4504.642 | 1.418 | 1.005 | 6750.0 | 6715.0 | 6767.0 | 4891.0 | 1.0 |
μ[1] | 3625.102 | 122.610 | 3403.476 | 3865.610 | 1.407 | 0.995 | 7597.0 | 7597.0 | 7622.0 | 4916.0 | 1.0 |
μ[2] | 3984.044 | 105.523 | 3785.491 | 4182.183 | 1.353 | 0.957 | 6082.0 | 6082.0 | 6082.0 | 4802.0 | 1.0 |
σ[0] | 78.531 | 59.788 | 0.019 | 184.598 | 0.819 | 0.692 | 5331.0 | 3737.0 | 4413.0 | 3198.0 | 1.0 |
σ[1] | 76.320 | 58.370 | 0.030 | 180.889 | 0.773 | 0.547 | 5699.0 | 5699.0 | 4607.0 | 3176.0 | 1.0 |
σ[2] | 85.882 | 64.475 | 0.266 | 201.367 | 0.970 | 0.686 | 4420.0 | 4420.0 | 3717.0 | 3351.0 | 1.0 |
az.plot_pair(output, var_names='μ', divergences=True);
az.plot_pair(output, var_names='σ', divergences=True);
Slower sampling¶
Once again, we have a few divergences, but they don't seem to say anything concrete as far as I can tell. These could be false positives; we could increase target_accept
, which is basically an inverse timestep for the simulation (shorter timesteps usually mean larger accuracy at the cost of more computational time).
with split_model:
trace = pm.sample(2000, tune=2000, chains=4, random_seed=1, target_accept=0.9)
output = az.from_pymc3(trace=trace,
prior=pm.sample_prior_predictive(2000 , var_names=predictive_var_names, random_seed=1),
posterior_predictive=pm.sample_posterior_predictive(trace, var_names=predictive_var_names, random_seed=1),
)
output
Auto-assigning NUTS sampler... Initializing NUTS using jitter+adapt_diag... Multiprocess sampling (4 chains in 2 jobs) NUTS: [helper, σ, μ]
Sampling 4 chains for 2_000 tune and 2_000 draw iterations (8_000 + 8_000 draws total) took 34 seconds.
-
-
- chain: 4
- draw: 2000
- race: 3
- replay: 138
-
-
chain(chain)int640 1 2 3
array([0, 1, 2, 3])
-
draw(draw)int640 1 2 3 4 ... 1996 1997 1998 1999
array([ 0, 1, 2, ..., 1997, 1998, 1999])
-
race(race)<U7'Terran' 'Protoss' 'Zerg'
array(['Terran', 'Protoss', 'Zerg'], dtype='<U7')
-
replay(replay)int648 325 54 346 138 ... 25 208 59 364
array([ 8, 325, 54, 346, 138, 405, 20, 129, 104, 46, 302, 408, 101, 219, 316, 126, 98, 231, 241, 385, 193, 198, 90, 329, 137, 200, 80, 355, 317, 33, 64, 213, 368, 49, 435, 134, 254, 330, 60, 39, 218, 109, 301, 133, 328, 181, 156, 395, 43, 249, 27, 153, 211, 420, 366, 186, 163, 63, 202, 45, 69, 31, 167, 177, 95, 151, 392, 387, 18, 286, 102, 290, 195, 428, 403, 97, 406, 412, 374, 263, 371, 41, 212, 52, 238, 345, 4, 117, 407, 56, 103, 118, 319, 57, 128, 294, 15, 273, 327, 281, 378, 121, 113, 284, 422, 389, 216, 58, 418, 309, 123, 116, 222, 122, 11, 361, 179, 255, 239, 225, 381, 424, 343, 287, 341, 184, 380, 196, 174, 306, 252, 148, 416, 391, 25, 208, 59, 364])
-
-
-
μ(chain, draw, race)float644.192e+03 3.663e+03 ... 3.958e+03
array([[[4191.82605585, 3662.56685124, 3976.78290968], [4257.97578329, 3613.94126582, 3981.76487183], [4298.83077594, 3669.01287593, 3980.01052311], ..., [4166.91919982, 3629.16545514, 3964.33087075], [4253.24812539, 3840.22261721, 3984.01357522], [4227.20217537, 3489.99729021, 3910.30300021]], [[4239.12879032, 3801.8761701 , 4067.56023194], [4217.78015886, 3898.22751246, 3973.50400805], [4484.11725818, 3779.99293822, 4016.31411823], ..., [4258.99732702, 3596.77273898, 4039.48381277], [4274.35096163, 3547.99306416, 4022.58218725], [4212.3466215 , 3574.92200423, 4175.91083343]], [[4108.92610165, 3552.9706627 , 3829.50707736], [4384.06283782, 3691.27109238, 4135.1039508 ], [4446.44987489, 3741.85060453, 4086.13686973], ..., [4335.40009373, 3731.15106938, 3962.55328323], [4307.456269 , 3749.69690735, 4120.08021688], [4190.8374136 , 3794.59313277, 4101.57910293]], [[4272.61598805, 3719.24049327, 4048.31502213], [4228.36765588, 3600.03935061, 3874.50140572], [4293.21198926, 3603.78459054, 3897.61309849], ..., [4389.92203176, 3410.26310882, 3951.09193896], [4282.64108689, 3725.26346768, 3981.54442129], [4266.90287295, 3532.39900759, 3958.19034856]]])
-
helper(chain, draw, replay)float64-0.3586 -1.85 ... 0.02351 -1.151
array([[[-3.58569651e-01, -1.84989802e+00, 2.29563873e-01, ..., 9.50090310e-01, 1.56085276e+00, 8.22590552e-01], [ 9.06258881e-01, -5.14902909e-01, -1.04868401e-01, ..., -1.62571690e+00, 6.70900325e-01, -1.51883191e-01], [-6.40631182e-01, 6.12057059e-01, 3.17121019e-01, ..., 1.36327897e+00, -1.11966799e+00, 1.72247767e-01], ..., [ 1.20412968e+00, -3.51638910e-01, 5.88442964e-01, ..., 9.28011498e-02, -5.64504440e-01, -7.09598482e-01], [-1.96695828e+00, -6.73132651e-01, 4.44071882e-02, ..., 3.86901656e-01, 4.41467830e-01, 3.20274641e-01], [ 1.42440822e+00, 6.00126327e-01, -3.94391564e-01, ..., -4.39375166e-01, -4.45562672e-01, -1.77496465e+00]], [[ 6.54007611e-01, -8.24447681e-01, 1.05125375e+00, ..., 1.17847687e+00, -6.69548837e-01, -1.15841741e+00], [ 1.37685327e+00, -9.87904032e-01, -1.09741324e-02, ..., 2.88743415e-01, -1.12658686e+00, -1.38173770e+00], [-3.18193693e-03, 2.41073061e-01, -1.22760501e-01, ..., -7.64371279e-02, -4.78951612e-01, 2.60067221e-01], ..., [-6.27546050e-01, 9.98509713e-01, -1.11550676e+00, ..., -2.26676217e-01, -1.38838943e-01, -9.93353582e-01], [ 5.13204543e-01, -1.58241396e+00, -1.08417712e-01, ..., 4.68034271e-01, 3.76969235e-01, -4.04638391e-01], [-1.76788541e-01, 9.32990055e-01, 4.28135341e-05, ..., 1.02912759e+00, -3.42634612e-02, -2.14568024e-01]], [[ 8.77482038e-01, 5.66063006e-01, -4.11165561e-01, ..., 5.01402289e-01, -3.55815931e-02, -8.21029200e-01], [-6.08466799e-01, -1.15355056e+00, 1.68031486e+00, ..., -4.10245950e-01, 4.59513709e-01, 8.49965623e-01], [-7.64269406e-01, -1.05121556e+00, 7.99346402e-01, ..., 1.65078612e-01, 1.01143791e+00, -1.84026189e-01], ..., [ 5.68352677e-01, -6.28746748e-01, 7.97804290e-01, ..., -3.36482081e-01, 1.62106906e-01, -2.15009903e+00], [ 1.59171993e+00, -3.55615603e-02, 2.37360931e+00, ..., -1.80848977e+00, 7.13096344e-01, -2.17528450e-01], [ 4.66494945e-01, -2.97893117e-01, 2.06685738e+00, ..., -9.37338897e-01, 1.00084739e+00, -3.86379260e-01]], [[ 1.67365693e+00, 7.21526403e-01, 1.06145479e+00, ..., 5.28478191e-01, 1.54626828e+00, -5.33971064e-01], [-1.32996411e+00, 4.15343513e-01, -1.47097832e+00, ..., -9.36878676e-01, -1.04900111e+00, 8.36467466e-01], [-1.11769228e-01, 2.42671312e+00, -8.76540192e-01, ..., -2.52865488e-01, -1.46753688e+00, -1.76933513e-01], ..., [ 8.56890908e-02, 2.39181911e-01, -8.68460585e-01, ..., -2.21486333e+00, -1.43487594e-01, -5.96289347e-01], [ 7.22463030e-02, -1.56334275e-01, 5.66600313e-01, ..., 2.37982042e+00, 3.10878601e-01, 1.35762265e+00], [-4.91907389e-02, 3.76161952e-01, -6.80508953e-01, ..., -1.94831032e+00, 2.35097946e-02, -1.15050624e+00]]])
-
σ(chain, draw, race)float64204.2 76.8 32.59 ... 31.32 94.91
array([[[2.04242226e+02, 7.68008837e+01, 3.25943353e+01], [4.18171668e+01, 7.45761818e+01, 1.19526142e+02], [8.97354023e+01, 6.62820028e+01, 3.64307641e+01], ..., [1.60576390e+02, 2.31497996e+00, 2.58897070e+01], [1.78390681e+01, 8.50597976e+00, 1.68203140e+02], [7.31420294e+01, 1.27927535e+02, 4.83013062e+00]], [[1.09849691e+02, 2.82949807e+01, 1.37658429e+02], [1.87770432e+02, 3.34908584e+01, 1.10070610e+02], [3.44636450e+01, 2.64758758e+02, 6.32728066e+01], ..., [1.08545694e+02, 1.90329689e+02, 7.53971341e+00], [9.69128164e+01, 1.47267519e+02, 4.71936992e+01], [3.24385154e+01, 1.87935685e+02, 1.02406167e+02]], [[2.49993613e+02, 7.63470122e+01, 1.28194539e+02], [1.22753266e+02, 1.00566005e+02, 1.26635005e+01], [3.82276057e+01, 5.74721016e+01, 1.87588800e+01], ..., [4.63901695e+01, 1.47619023e+00, 1.38726849e+02], [2.26734894e+02, 5.42758590e+01, 2.25301282e+01], [5.77676739e+01, 1.64411686e+02, 8.20206053e+01]], [[5.11084577e+01, 2.58581172e+02, 6.86281538e+01], [1.08308154e+02, 2.44693323e-01, 1.58790827e+02], [1.99750479e+02, 1.51796727e+00, 9.00794021e+01], ..., [6.66429227e+01, 1.34972643e+02, 9.20364312e+01], [1.06224961e+02, 3.81711978e+01, 9.23637668e+01], [4.57938897e+01, 3.13248774e+01, 9.49120778e+01]]])
-
MMR(chain, draw, replay)float643.635e+03 3.52e+03 ... 3.849e+03
array([[[3635.02838514, 3520.49304836, 3984.26539153, ..., 4385.87461587, 4510.6180973 , 4003.59470195], [3681.5265929 , 3575.54177283, 3969.2303564 , ..., 4189.99290857, 4286.03093408, 3963.61085993], [3626.55055818, 3709.58124363, 3991.56348415, ..., 4421.16516306, 4198.35691828, 3986.28564086], ..., [3631.95299122, 3628.35141811, 3979.56548664, ..., 4181.82087345, 4076.27311465, 3945.95957399], [3823.49170987, 3834.49696451, 3991.48300371, ..., 4260.15009037, 4261.12350006, 4037.88477541], [3672.21832284, 3566.76997226, 3908.39803744, ..., 4195.06538405, 4194.6128173 , 3901.72968909]], [[3820.3813028 , 3778.54843892, 4212.27417155, ..., 4368.58411007, 4165.57905776, 3908.09431138], [3944.33951023, 3865.14175844, 3972.2960786 , ..., 4271.99763448, 4006.24045846, 3821.41529706], [3779.15049255, 3843.81914225, 4008.54671677, ..., 4481.48295614, 4467.61083985, 4032.76930122], ..., [3477.33209465, 3786.8187819 , 4031.07321149, ..., 4234.3925996 , 4243.92695754, 4031.99421144], [3623.57142381, 3314.95488652, 4017.46555434, ..., 4319.70948106, 4310.88411195, 4003.48580472], [3541.69712857, 3750.26412962, 4175.9152178 , ..., 4245.72999278, 4211.23516568, 4153.93774455]], [[3619.96379451, 3596.18788193, 3776.79789789, ..., 4234.27347158, 4100.03093064, 3724.25561768], [3630.08001753, 3575.26312196, 4156.3826188 , ..., 4333.70380745, 4440.4696466 , 4145.86749087], [3697.92643555, 3681.43503719, 4101.13171294, ..., 4452.76043496, 4485.11472444, 4082.68474453], ..., [3731.99006605, 3730.22291957, 4073.23015874, ..., 4319.79063297, 4342.92026057, 3664.27681853], [3836.08887366, 3747.76677311, 4173.55793881, ..., 3897.40853377, 4469.14009285, 4115.17927302], [3871.290353 , 3745.61602321, 4271.10399599, ..., 4136.6895259 , 4248.65403921, 4069.88804217]], [[4152.01666383, 3905.81363602, 4121.16070472, ..., 4299.62569332, 4351.64337494, 4011.66957382], [3599.71391728, 3600.1409824 , 3640.92354229, ..., 4126.8960557 , 4114.7522814 , 4007.32476636], [3603.61492851, 3607.46826165, 3818.6548821 , ..., 4242.70198684, 4000.07079332, 3881.67503342], ..., [3421.82879184, 3442.54612337, 3871.1619261 , ..., 4242.31706615, 4380.35959915, 3896.21159555], [3728.0211956 , 3719.29600116, 4033.87776046, ..., 4535.43741766, 4315.66415412, 4106.93956281], [3530.85811372, 3544.18223464, 3893.60182986, ..., 4177.68216515, 4267.9794779 , 3848.99341069]]])
-
MMR_diff(chain, draw, replay)float64-113.0 -425.5 ... 331.0 -391.0
array([[[-112.97161486, -425.50695164, 182.26539153, ..., 408.87461587, 573.6180973 , -236.40529805], [ -66.4734071 , -370.45822717, 167.2303564 , ..., 212.99290857, 349.03093408, -276.38914007], [-121.44944182, -236.41875637, 189.56348415, ..., 444.16516306, 261.35691828, -253.71435914], ..., [-116.04700878, -317.64858189, 177.56548664, ..., 204.82087345, 139.27311465, -294.04042601], [ 75.49170987, -111.50303549, 189.48300371, ..., 283.15009037, 324.12350006, -202.11522459], [ -75.78167716, -379.23002774, 106.39803744, ..., 218.06538405, 257.6128173 , -338.27031091]], [[ 72.3813028 , -167.45156108, 410.27417155, ..., 391.58411007, 228.57905776, -331.90568862], [ 196.33951023, -80.85824156, 170.2960786 , ..., 294.99763448, 69.24045846, -418.58470294], [ 31.15049255, -102.18085775, 206.54671677, ..., 504.48295614, 530.61083985, -207.23069878], ..., [-270.66790535, -159.1812181 , 229.07321149, ..., 257.3925996 , 306.92695754, -208.00578856], [-124.42857619, -631.04511348, 215.46555434, ..., 342.70948106, 373.88411195, -236.51419528], [-206.30287143, -195.73587038, 373.9152178 , ..., 268.72999278, 274.23516568, -86.06225545]], [[-128.03620549, -349.81211807, -25.20210211, ..., 257.27347158, 163.03093064, -515.74438232], [-117.91998247, -370.73687804, 354.3826188 , ..., 356.70380745, 503.4696466 , -94.13250913], [ -50.07356445, -264.56496281, 299.13171294, ..., 475.76043496, 548.11472444, -157.31525547], ..., [ -16.00993395, -215.77708043, 271.23015874, ..., 342.79063297, 405.92026057, -575.72318147], [ 88.08887366, -198.23322689, 371.55793881, ..., -79.59146623, 532.14009285, -124.82072698], [ 123.290353 , -200.38397679, 469.10399599, ..., 159.6895259 , 311.65403921, -170.11195783]], [[ 404.01666383, -40.18636398, 319.16070472, ..., 322.62569332, 414.64337494, -228.33042618], [-148.28608272, -345.8590176 , -161.07645771, ..., 149.8960557 , 177.7522814 , -232.67523364], [-144.38507149, -338.53173835, 16.6548821 , ..., 265.70198684, 63.07079332, -358.32496658], ..., [-326.17120816, -503.45387663, 69.1619261 , ..., 265.31706615, 443.35959915, -343.78840445], [ -19.9788044 , -226.70399884, 231.87776046, ..., 558.43741766, 378.66415412, -133.06043719], [-217.14188628, -401.81776536, 91.60182986, ..., 200.68216515, 330.9794779 , -391.00658931]]])
-
winrate(chain, draw, replay)float640.4266 0.2472 ... 0.7039 0.2644
array([[[0.4266338 , 0.24724307, 0.61701816, ..., 0.74456852, 0.81771078, 0.35010631], [0.45662615, 0.2750138 , 0.60768 , ..., 0.63583029, 0.71366787, 0.32669033], [0.42121652, 0.3500983 , 0.62152047, ..., 0.7617306 , 0.6645996 , 0.33987251], ..., [0.42466653, 0.30340175, 0.61410799, ..., 0.63086488, 0.59010958, 0.31661356], [0.54922247, 0.42757404, 0.62147094, ..., 0.67718895, 0.70016698, 0.37078528], [0.4505897 , 0.2704613 , 0.56915351, ..., 0.63889797, 0.66241233, 0.29212013]], [[0.54720674, 0.39218202, 0.74526436, ..., 0.73586936, 0.64522013, 0.29557572], [0.62568207, 0.44730353, 0.60959075, ..., 0.68392811, 0.54516971, 0.25062949], [0.52036562, 0.4335544 , 0.63191587, ..., 0.78918453, 0.80033389, 0.36766795], ..., [0.3299917 , 0.39735231, 0.64551605, ..., 0.66228347, 0.69063659, 0.36719657], [0.4193173 , 0.16095004, 0.63732707, ..., 0.71027596, 0.72676986, 0.35004148], [0.36823254, 0.37468792, 0.72678602, ..., 0.66888622, 0.67206874, 0.44393964]], [[0.41702061, 0.28591462, 0.48352018, ..., 0.66221375, 0.6050573 , 0.20595485], [0.4234696 , 0.27486845, 0.71652076, ..., 0.71775279, 0.78874307, 0.43873325], [0.46729148, 0.33353189, 0.68626179, ..., 0.77640925, 0.8075522 , 0.39852206], ..., [0.48952874, 0.36248454, 0.67033349, ..., 0.71031966, 0.74309555, 0.18146961], [0.55736901, 0.37315816, 0.72555955, ..., 0.44812312, 0.80097255, 0.41906748], [0.57995735, 0.37184275, 0.77337113, ..., 0.60296614, 0.69327302, 0.39052391]], [[0.74214352, 0.47373655, 0.69743382, ..., 0.69934358, 0.74742865, 0.35492881], [0.40419849, 0.28803111, 0.39616541, ..., 0.59681559, 0.61422381, 0.35233025], [0.40665901, 0.2919787 , 0.51089296, ..., 0.66712911, 0.54116397, 0.28138867], ..., [0.29870941, 0.2112638 , 0.54511876, ..., 0.66690541, 0.76134783, 0.28914344], [0.48693397, 0.35590376, 0.64719344, ..., 0.81171504, 0.72924646, 0.41382809], [0.3616597 , 0.25895906, 0.55963552, ..., 0.62833948, 0.70391943, 0.26442445]]])
-
-
- created_at :
- 2020-06-20T14:25:20.065376
- arviz_version :
- 0.8.3
- inference_library :
- pymc3
- inference_library_version :
- 3.9.1
- sampling_time :
- 34.35768222808838
- tuning_steps :
- 2000
xarray.Dataset
-
-
-
- chain: 4
- draw: 2000
- race: 3
- replay: 138
-
-
chain(chain)int640 1 2 3
array([0, 1, 2, 3])
-
draw(draw)int640 1 2 3 4 ... 1996 1997 1998 1999
array([ 0, 1, 2, ..., 1997, 1998, 1999])
-
replay(replay)int648 325 54 346 138 ... 25 208 59 364
array([ 8, 325, 54, 346, 138, 405, 20, 129, 104, 46, 302, 408, 101, 219, 316, 126, 98, 231, 241, 385, 193, 198, 90, 329, 137, 200, 80, 355, 317, 33, 64, 213, 368, 49, 435, 134, 254, 330, 60, 39, 218, 109, 301, 133, 328, 181, 156, 395, 43, 249, 27, 153, 211, 420, 366, 186, 163, 63, 202, 45, 69, 31, 167, 177, 95, 151, 392, 387, 18, 286, 102, 290, 195, 428, 403, 97, 406, 412, 374, 263, 371, 41, 212, 52, 238, 345, 4, 117, 407, 56, 103, 118, 319, 57, 128, 294, 15, 273, 327, 281, 378, 121, 113, 284, 422, 389, 216, 58, 418, 309, 123, 116, 222, 122, 11, 361, 179, 255, 239, 225, 381, 424, 343, 287, 341, 184, 380, 196, 174, 306, 252, 148, 416, 391, 25, 208, 59, 364])
-
race(race)<U7'Terran' 'Protoss' 'Zerg'
array(['Terran', 'Protoss', 'Zerg'], dtype='<U7')
-
-
-
win(chain, draw, replay)int640 0 1 1 0 0 1 0 ... 1 0 1 0 0 1 1 0
array([[[0, 0, 1, ..., 1, 1, 0], [1, 1, 1, ..., 1, 0, 1], [0, 0, 1, ..., 1, 0, 1], ..., [1, 0, 0, ..., 1, 1, 1], [0, 1, 1, ..., 1, 1, 0], [0, 0, 1, ..., 1, 1, 1]], [[0, 0, 0, ..., 1, 1, 0], [0, 0, 1, ..., 1, 1, 0], [1, 0, 1, ..., 1, 1, 0], ..., [1, 0, 1, ..., 0, 0, 1], [1, 0, 1, ..., 1, 1, 0], [1, 0, 1, ..., 0, 1, 0]], [[1, 1, 1, ..., 0, 0, 0], [1, 0, 1, ..., 1, 1, 0], [1, 1, 0, ..., 1, 1, 1], ..., [0, 0, 0, ..., 1, 1, 0], [1, 1, 1, ..., 0, 1, 0], [0, 1, 1, ..., 1, 1, 0]], [[1, 1, 1, ..., 1, 1, 0], [0, 0, 0, ..., 1, 0, 0], [1, 1, 0, ..., 1, 0, 1], ..., [0, 0, 1, ..., 0, 1, 1], [0, 0, 0, ..., 1, 0, 0], [1, 0, 0, ..., 1, 1, 0]]])
-
μ(chain, draw, race)float644.192e+03 3.663e+03 ... 3.958e+03
array([[[4191.82605585, 3662.56685124, 3976.78290968], [4257.97578329, 3613.94126582, 3981.76487183], [4298.83077594, 3669.01287593, 3980.01052311], ..., [4166.91919982, 3629.16545514, 3964.33087075], [4253.24812539, 3840.22261721, 3984.01357522], [4227.20217537, 3489.99729021, 3910.30300021]], [[4239.12879032, 3801.8761701 , 4067.56023194], [4217.78015886, 3898.22751246, 3973.50400805], [4484.11725818, 3779.99293822, 4016.31411823], ..., [4258.99732702, 3596.77273898, 4039.48381277], [4274.35096163, 3547.99306416, 4022.58218725], [4212.3466215 , 3574.92200423, 4175.91083343]], [[4108.92610165, 3552.9706627 , 3829.50707736], [4384.06283782, 3691.27109238, 4135.1039508 ], [4446.44987489, 3741.85060453, 4086.13686973], ..., [4335.40009373, 3731.15106938, 3962.55328323], [4307.456269 , 3749.69690735, 4120.08021688], [4190.8374136 , 3794.59313277, 4101.57910293]], [[4272.61598805, 3719.24049327, 4048.31502213], [4228.36765588, 3600.03935061, 3874.50140572], [4293.21198926, 3603.78459054, 3897.61309849], ..., [4389.92203176, 3410.26310882, 3951.09193896], [4282.64108689, 3725.26346768, 3981.54442129], [4266.90287295, 3532.39900759, 3958.19034856]]])
-
σ(chain, draw, race)float64204.2 76.8 32.59 ... 31.32 94.91
array([[[2.04242226e+02, 7.68008837e+01, 3.25943353e+01], [4.18171668e+01, 7.45761818e+01, 1.19526142e+02], [8.97354023e+01, 6.62820028e+01, 3.64307641e+01], ..., [1.60576390e+02, 2.31497996e+00, 2.58897070e+01], [1.78390681e+01, 8.50597976e+00, 1.68203140e+02], [7.31420294e+01, 1.27927535e+02, 4.83013062e+00]], [[1.09849691e+02, 2.82949807e+01, 1.37658429e+02], [1.87770432e+02, 3.34908584e+01, 1.10070610e+02], [3.44636450e+01, 2.64758758e+02, 6.32728066e+01], ..., [1.08545694e+02, 1.90329689e+02, 7.53971341e+00], [9.69128164e+01, 1.47267519e+02, 4.71936992e+01], [3.24385154e+01, 1.87935685e+02, 1.02406167e+02]], [[2.49993613e+02, 7.63470122e+01, 1.28194539e+02], [1.22753266e+02, 1.00566005e+02, 1.26635005e+01], [3.82276057e+01, 5.74721016e+01, 1.87588800e+01], ..., [4.63901695e+01, 1.47619023e+00, 1.38726849e+02], [2.26734894e+02, 5.42758590e+01, 2.25301282e+01], [5.77676739e+01, 1.64411686e+02, 8.20206053e+01]], [[5.11084577e+01, 2.58581172e+02, 6.86281538e+01], [1.08308154e+02, 2.44693323e-01, 1.58790827e+02], [1.99750479e+02, 1.51796727e+00, 9.00794021e+01], ..., [6.66429227e+01, 1.34972643e+02, 9.20364312e+01], [1.06224961e+02, 3.81711978e+01, 9.23637668e+01], [4.57938897e+01, 3.13248774e+01, 9.49120778e+01]]])
-
winrate(chain, draw, replay)float640.4266 0.2472 ... 0.7039 0.2644
array([[[0.4266338 , 0.24724307, 0.61701816, ..., 0.74456852, 0.81771078, 0.35010631], [0.45662615, 0.2750138 , 0.60768 , ..., 0.63583029, 0.71366787, 0.32669033], [0.42121652, 0.3500983 , 0.62152047, ..., 0.7617306 , 0.6645996 , 0.33987251], ..., [0.42466653, 0.30340175, 0.61410799, ..., 0.63086488, 0.59010958, 0.31661356], [0.54922247, 0.42757404, 0.62147094, ..., 0.67718895, 0.70016698, 0.37078528], [0.4505897 , 0.2704613 , 0.56915351, ..., 0.63889797, 0.66241233, 0.29212013]], [[0.54720674, 0.39218202, 0.74526436, ..., 0.73586936, 0.64522013, 0.29557572], [0.62568207, 0.44730353, 0.60959075, ..., 0.68392811, 0.54516971, 0.25062949], [0.52036562, 0.4335544 , 0.63191587, ..., 0.78918453, 0.80033389, 0.36766795], ..., [0.3299917 , 0.39735231, 0.64551605, ..., 0.66228347, 0.69063659, 0.36719657], [0.4193173 , 0.16095004, 0.63732707, ..., 0.71027596, 0.72676986, 0.35004148], [0.36823254, 0.37468792, 0.72678602, ..., 0.66888622, 0.67206874, 0.44393964]], [[0.41702061, 0.28591462, 0.48352018, ..., 0.66221375, 0.6050573 , 0.20595485], [0.4234696 , 0.27486845, 0.71652076, ..., 0.71775279, 0.78874307, 0.43873325], [0.46729148, 0.33353189, 0.68626179, ..., 0.77640925, 0.8075522 , 0.39852206], ..., [0.48952874, 0.36248454, 0.67033349, ..., 0.71031966, 0.74309555, 0.18146961], [0.55736901, 0.37315816, 0.72555955, ..., 0.44812312, 0.80097255, 0.41906748], [0.57995735, 0.37184275, 0.77337113, ..., 0.60296614, 0.69327302, 0.39052391]], [[0.74214352, 0.47373655, 0.69743382, ..., 0.69934358, 0.74742865, 0.35492881], [0.40419849, 0.28803111, 0.39616541, ..., 0.59681559, 0.61422381, 0.35233025], [0.40665901, 0.2919787 , 0.51089296, ..., 0.66712911, 0.54116397, 0.28138867], ..., [0.29870941, 0.2112638 , 0.54511876, ..., 0.66690541, 0.76134783, 0.28914344], [0.48693397, 0.35590376, 0.64719344, ..., 0.81171504, 0.72924646, 0.41382809], [0.3616597 , 0.25895906, 0.55963552, ..., 0.62833948, 0.70391943, 0.26442445]]])
-
-
- created_at :
- 2020-06-20T14:25:20.600612
- arviz_version :
- 0.8.3
- inference_library :
- pymc3
- inference_library_version :
- 3.9.1
xarray.Dataset
-
-
-
- chain: 4
- draw: 2000
- replay: 138
-
-
chain(chain)int640 1 2 3
array([0, 1, 2, 3])
-
draw(draw)int640 1 2 3 4 ... 1996 1997 1998 1999
array([ 0, 1, 2, ..., 1997, 1998, 1999])
-
replay(replay)int648 325 54 346 138 ... 25 208 59 364
array([ 8, 325, 54, 346, 138, 405, 20, 129, 104, 46, 302, 408, 101, 219, 316, 126, 98, 231, 241, 385, 193, 198, 90, 329, 137, 200, 80, 355, 317, 33, 64, 213, 368, 49, 435, 134, 254, 330, 60, 39, 218, 109, 301, 133, 328, 181, 156, 395, 43, 249, 27, 153, 211, 420, 366, 186, 163, 63, 202, 45, 69, 31, 167, 177, 95, 151, 392, 387, 18, 286, 102, 290, 195, 428, 403, 97, 406, 412, 374, 263, 371, 41, 212, 52, 238, 345, 4, 117, 407, 56, 103, 118, 319, 57, 128, 294, 15, 273, 327, 281, 378, 121, 113, 284, 422, 389, 216, 58, 418, 309, 123, 116, 222, 122, 11, 361, 179, 255, 239, 225, 381, 424, 343, 287, 341, 184, 380, 196, 174, 306, 252, 148, 416, 391, 25, 208, 59, 364])
-
-
-
win(chain, draw, replay)float64-0.8518 -0.284 ... -0.3511 -1.33
array([[[-0.85182925, -0.28401291, -0.9597677 , ..., -0.2949504 , -0.20124657, -1.04951843], [-0.78389028, -0.32160266, -0.93567745, ..., -0.45282359, -0.33733759, -1.11874255], [-0.86460828, -0.43093415, -0.9715933 , ..., -0.27216232, -0.40857053, -1.07918469], ..., [-0.85645105, -0.36154643, -0.95219772, ..., -0.46066357, -0.52744702, -1.1500733 ], [-0.5992517 , -0.55787188, -0.97146243, ..., -0.38980495, -0.35643643, -0.99213215], [-0.79719812, -0.31534286, -0.84200343, ..., -0.44801051, -0.41186706, -1.23059016]], [[-0.6029286 , -0.49787982, -1.36752899, ..., -0.30670267, -0.43816374, -1.21883022], [-0.46891291, -0.59294631, -0.94055973, ..., -0.37990247, -0.60665814, -1.38377958], [-0.6532236 , -0.56837423, -0.99944377, ..., -0.23675511, -0.22272627, -1.00057507], ..., [-1.10868777, -0.50642252, -1.03709221, ..., -0.41206162, -0.37014152, -1.00185797], [-0.86912737, -0.17548503, -1.01425387, ..., -0.3421017 , -0.31914541, -1.04970362], [-0.99904063, -0.46950443, -1.2975 , ..., -0.40214131, -0.39739466, -0.81206668]], [[-0.87461964, -0.33675275, -0.66071906, ..., -0.4121669 , -0.50243211, -1.58009829], [-0.85927356, -0.32140219, -1.26061639, ..., -0.33163007, -0.23731465, -0.82386367], [-0.76080206, -0.40576299, -1.15919636, ..., -0.25307552, -0.21374759, -0.91999242], ..., [-0.71431211, -0.45017676, -1.10967372, ..., -0.34204019, -0.29693065, -1.70666705], [-0.58452777, -0.46706102, -1.29302097, ..., -0.80268726, -0.2219286 , -0.86972333], [-0.54480071, -0.46496474, -1.48444153, ..., -0.50589424, -0.36633139, -0.94026608]], [[-0.29821263, -0.64195333, -1.19545525, ..., -0.35761313, -0.29111643, -1.03583805], [-0.90584921, -0.33972107, -0.50445498, ..., -0.51614711, -0.48739591, -1.04318632], [-0.89978025, -0.3452811 , -0.71517392, ..., -0.40477168, -0.61403296, -1.26801841], ..., [-1.20828404, -0.23732337, -0.7877189 , ..., -0.40510705, -0.27266496, -1.2408324 ], [-0.71962675, -0.43990713, -1.04183536, ..., -0.20860593, -0.31574353, -0.88230464], [-1.01705156, -0.2996994 , -0.82015253, ..., -0.46467469, -0.35109138, -1.33019969]]])
-
-
- created_at :
- 2020-06-20T14:25:20.597504
- arviz_version :
- 0.8.3
- inference_library :
- pymc3
- inference_library_version :
- 3.9.1
xarray.Dataset
-
-
-
- chain: 4
- draw: 2000
-
-
chain(chain)int640 1 2 3
array([0, 1, 2, 3])
-
draw(draw)int640 1 2 3 4 ... 1996 1997 1998 1999
array([ 0, 1, 2, ..., 1997, 1998, 1999])
-
-
-
step_size_bar(chain, draw)float640.294 0.294 0.294 ... 0.2907 0.2907
array([[0.29398781, 0.29398781, 0.29398781, ..., 0.29398781, 0.29398781, 0.29398781], [0.32832862, 0.32832862, 0.32832862, ..., 0.32832862, 0.32832862, 0.32832862], [0.3100242 , 0.3100242 , 0.3100242 , ..., 0.3100242 , 0.3100242 , 0.3100242 ], [0.29071728, 0.29071728, 0.29071728, ..., 0.29071728, 0.29071728, 0.29071728]])
-
energy_error(chain, draw)float64-1.165 0.1154 ... 0.2452 -0.1032
array([[-1.16545649, 0.11543079, 0.05534909, ..., -0.02356721, 0.01416864, 0.22002702], [-0.27486769, 0.46115983, 0.41228911, ..., -0.07802792, 0.1259835 , -0.33367031], [-1.04578413, 0.16843381, 0.26006434, ..., -0.22386483, -0.24002094, 0.22646569], [-0.079549 , 0.96347628, 0.0941145 , ..., 0.06526721, 0.24516089, -0.10319888]])
-
step_size(chain, draw)float640.2628 0.2628 ... 0.3427 0.3427
array([[0.26275736, 0.26275736, 0.26275736, ..., 0.26275736, 0.26275736, 0.26275736], [0.32445057, 0.32445057, 0.32445057, ..., 0.32445057, 0.32445057, 0.32445057], [0.33329653, 0.33329653, 0.33329653, ..., 0.33329653, 0.33329653, 0.33329653], [0.34267612, 0.34267612, 0.34267612, ..., 0.34267612, 0.34267612, 0.34267612]])
-
max_energy_error(chain, draw)float64-1.165 0.2965 ... 0.5785 -0.4783
array([[-1.16545649, 0.2964833 , -0.21895137, ..., 0.48664513, 0.40310784, -0.67851845], [ 0.67112042, 0.55722514, 0.41228911, ..., -0.28722158, 0.20634744, -0.3871516 ], [ 1.49793689, 0.32854877, 0.92926063, ..., 1.32199749, 7.60500613, 1.06863903], [-0.32793756, 3.07300588, 3.39102021, ..., 0.42629728, 0.57845417, -0.47834955]])
-
diverging(chain, draw)boolFalse False False ... False False
array([[False, False, False, ..., False, False, False], [False, False, False, ..., False, False, False], [False, False, False, ..., False, False, False], [False, False, False, ..., False, False, False]])
-
depth(chain, draw)int644 4 4 4 4 4 4 4 ... 4 4 4 4 4 4 4 4
array([[4, 4, 4, ..., 4, 4, 4], [4, 4, 4, ..., 4, 4, 4], [4, 4, 4, ..., 4, 4, 4], [4, 4, 4, ..., 4, 4, 4]])
-
mean_tree_accept(chain, draw)float640.9887 0.9228 ... 0.7224 0.9945
array([[0.9886637 , 0.92280831, 0.97142841, ..., 0.8910868 , 0.88488932, 0.93672292], [0.90066886, 0.80011353, 0.90867509, ..., 0.99015461, 0.92949944, 0.99230292], [0.95258493, 0.82788819, 0.61444807, ..., 0.95654644, 0.9694625 , 0.6588462 ], [0.99969745, 0.35548557, 0.94576827, ..., 0.83801927, 0.72241791, 0.99445689]])
-
energy(chain, draw)float64372.8 372.4 379.1 ... 383.8 366.4
array([[372.76445605, 372.4047956 , 379.13879775, ..., 388.4625056 , 379.20274373, 363.47543194], [364.17887204, 357.96746207, 377.37892682, ..., 370.05542217, 371.67089236, 364.45527508], [376.65409707, 351.05076247, 380.10629174, ..., 389.53229808, 377.4382392 , 382.51445418], [390.21491721, 407.90760972, 391.36807296, ..., 382.1347993 , 383.81306147, 366.41006492]])
-
lp(chain, draw)float64-301.0 -307.2 ... -306.8 -301.5
array([[-300.95481382, -307.18827597, -308.94357562, ..., -307.24068004, -299.47974489, -306.848432 ], [-288.68782192, -302.46946337, -313.96012231, ..., -305.47765428, -306.96534549, -299.15644208], [-292.01283788, -290.82367657, -296.9325567 , ..., -309.11026093, -303.66544734, -307.93238498], [-311.48221322, -321.23575274, -318.54558226, ..., -303.81632684, -306.83830943, -301.48182129]])
-
tree_size(chain, draw)float6415.0 15.0 15.0 ... 15.0 15.0 15.0
array([[15., 15., 15., ..., 15., 15., 15.], [15., 15., 15., ..., 15., 15., 15.], [15., 15., 15., ..., 15., 15., 15.], [15., 15., 15., ..., 15., 15., 15.]])
-
-
- created_at :
- 2020-06-20T14:25:20.072749
- arviz_version :
- 0.8.3
- inference_library :
- pymc3
- inference_library_version :
- 3.9.1
- sampling_time :
- 34.35768222808838
- tuning_steps :
- 2000
xarray.Dataset
-
-
-
- chain: 1
- draw: 2000
- race: 3
- replay: 138
-
-
chain(chain)int640
array([0])
-
draw(draw)int640 1 2 3 4 ... 1996 1997 1998 1999
array([ 0, 1, 2, ..., 1997, 1998, 1999])
-
race(race)<U7'Terran' 'Protoss' 'Zerg'
array(['Terran', 'Protoss', 'Zerg'], dtype='<U7')
-
replay(replay)int648 325 54 346 138 ... 25 208 59 364
array([ 8, 325, 54, 346, 138, 405, 20, 129, 104, 46, 302, 408, 101, 219, 316, 126, 98, 231, 241, 385, 193, 198, 90, 329, 137, 200, 80, 355, 317, 33, 64, 213, 368, 49, 435, 134, 254, 330, 60, 39, 218, 109, 301, 133, 328, 181, 156, 395, 43, 249, 27, 153, 211, 420, 366, 186, 163, 63, 202, 45, 69, 31, 167, 177, 95, 151, 392, 387, 18, 286, 102, 290, 195, 428, 403, 97, 406, 412, 374, 263, 371, 41, 212, 52, 238, 345, 4, 117, 407, 56, 103, 118, 319, 57, 128, 294, 15, 273, 327, 281, 378, 121, 113, 284, 422, 389, 216, 58, 418, 309, 123, 116, 222, 122, 11, 361, 179, 255, 239, 225, 381, 424, 343, 287, 341, 184, 380, 196, 174, 306, 252, 148, 416, 391, 25, 208, 59, 364])
-
-
-
μ(chain, draw, race)float644.487e+03 3.816e+03 ... 4.079e+03
array([[[4487.3036091 , 3816.4730759 , 3841.54847432], [3678.10941335, 4259.6222888 , 3309.53839094], [4523.44352926, 3771.63792973, 4095.71172882], ..., [3793.7352304 , 3936.80991215, 3897.05804332], [4287.27113474, 4067.38137671, 3856.85673442], [3792.63645476, 3763.38063103, 4079.25539045]]])
-
σ(chain, draw, race)float6410.45 53.54 169.3 ... 210.5 146.2
array([[[ 10.45030045, 53.53690568, 169.32540234], [ 23.57617238, 10.93212404, 83.18141646], [125.6331917 , 94.68059879, 168.58509729], ..., [ 54.01739286, 112.35908554, 107.02562404], [144.26356103, 77.82838645, 9.93539497], [ 81.81960096, 210.54208204, 146.24259794]]])
-
winrate(chain, draw, replay)float640.4452 0.4443 ... 0.4049 0.3675
array([[[0.44516586, 0.44432904, 0.49543905, ..., 0.79673329, 0.80140205, 0.28212387], [0.7972168 , 0.68737573, 0.21071926, ..., 0.31760374, 0.3228919 , 0.08274033], [0.46947403, 0.43159697, 0.78947532, ..., 0.82439068, 0.80940984, 0.32718542], ..., [0.74439685, 0.57396175, 0.48594935, ..., 0.37104134, 0.37754386, 0.3333573 ], [0.70701564, 0.56655281, 0.54098671, ..., 0.57389202, 0.76269576, 0.26823939], [0.59530994, 0.28697282, 0.72372083, ..., 0.36763267, 0.40492739, 0.36749525]]])
-
-
- created_at :
- 2020-06-20T14:25:20.604059
- arviz_version :
- 0.8.3
- inference_library :
- pymc3
- inference_library_version :
- 3.9.1
xarray.Dataset
-
-
-
- chain: 1
- draw: 2000
- replay: 138
-
-
chain(chain)int640
array([0])
-
draw(draw)int640 1 2 3 4 ... 1996 1997 1998 1999
array([ 0, 1, 2, ..., 1997, 1998, 1999])
-
replay(replay)int648 325 54 346 138 ... 25 208 59 364
array([ 8, 325, 54, 346, 138, 405, 20, 129, 104, 46, 302, 408, 101, 219, 316, 126, 98, 231, 241, 385, 193, 198, 90, 329, 137, 200, 80, 355, 317, 33, 64, 213, 368, 49, 435, 134, 254, 330, 60, 39, 218, 109, 301, 133, 328, 181, 156, 395, 43, 249, 27, 153, 211, 420, 366, 186, 163, 63, 202, 45, 69, 31, 167, 177, 95, 151, 392, 387, 18, 286, 102, 290, 195, 428, 403, 97, 406, 412, 374, 263, 371, 41, 212, 52, 238, 345, 4, 117, 407, 56, 103, 118, 319, 57, 128, 294, 15, 273, 327, 281, 378, 121, 113, 284, 422, 389, 216, 58, 418, 309, 123, 116, 222, 122, 11, 361, 179, 255, 239, 225, 381, 424, 343, 287, 341, 184, 380, 196, 174, 306, 252, 148, 416, 391, 25, 208, 59, 364])
-
-
-
win(chain, draw, replay)int641 1 0 0 1 1 1 0 ... 0 0 0 1 1 0 0 0
array([[[1, 1, 0, ..., 1, 1, 0], [1, 1, 0, ..., 0, 0, 0], [1, 0, 0, ..., 1, 1, 0], ..., [1, 1, 0, ..., 0, 0, 0], [1, 1, 0, ..., 1, 0, 0], [0, 0, 0, ..., 0, 0, 0]]])
-
-
- created_at :
- 2020-06-20T14:25:20.605812
- arviz_version :
- 0.8.3
- inference_library :
- pymc3
- inference_library_version :
- 3.9.1
xarray.Dataset
-
-
-
- replay: 138
-
-
replay(replay)int648 325 54 346 138 ... 25 208 59 364
array([ 8, 325, 54, 346, 138, 405, 20, 129, 104, 46, 302, 408, 101, 219, 316, 126, 98, 231, 241, 385, 193, 198, 90, 329, 137, 200, 80, 355, 317, 33, 64, 213, 368, 49, 435, 134, 254, 330, 60, 39, 218, 109, 301, 133, 328, 181, 156, 395, 43, 249, 27, 153, 211, 420, 366, 186, 163, 63, 202, 45, 69, 31, 167, 177, 95, 151, 392, 387, 18, 286, 102, 290, 195, 428, 403, 97, 406, 412, 374, 263, 371, 41, 212, 52, 238, 345, 4, 117, 407, 56, 103, 118, 319, 57, 128, 294, 15, 273, 327, 281, 378, 121, 113, 284, 422, 389, 216, 58, 418, 309, 123, 116, 222, 122, 11, 361, 179, 255, 239, 225, 381, 424, 343, 287, 341, 184, 380, 196, 174, 306, 252, 148, 416, 391, 25, 208, 59, 364])
-
-
-
win(replay)float641.0 0.0 0.0 1.0 ... 1.0 1.0 1.0 1.0
array([1., 0., 0., 1., 1., 0., 1., 0., 1., 0., 0., 1., 0., 0., 0., 0., 1., 1., 0., 0., 0., 1., 0., 0., 1., 0., 0., 1., 1., 1., 1., 1., 1., 0., 0., 0., 1., 0., 1., 1., 1., 1., 1., 1., 1., 0., 1., 0., 1., 0., 1., 0., 1., 0., 1., 1., 1., 1., 0., 0., 1., 0., 0., 1., 0., 0., 0., 1., 1., 1., 0., 1., 1., 0., 0., 1., 1., 1., 1., 0., 1., 0., 1., 1., 0., 0., 1., 0., 1., 0., 1., 1., 0., 1., 0., 0., 0., 0., 1., 0., 0., 0., 1., 0., 1., 0., 1., 0., 1., 0., 1., 0., 1., 0., 0., 0., 1., 1., 1., 0., 0., 0., 0., 1., 1., 0., 1., 0., 0., 1., 1., 1., 1., 0., 1., 1., 1., 1.])
-
-
- created_at :
- 2020-06-20T14:25:20.606613
- arviz_version :
- 0.8.3
- inference_library :
- pymc3
- inference_library_version :
- 3.9.1
xarray.Dataset
-
-
-
- replay: 138
-
-
replay(replay)int648 325 54 346 138 ... 25 208 59 364
array([ 8, 325, 54, 346, 138, 405, 20, 129, 104, 46, 302, 408, 101, 219, 316, 126, 98, 231, 241, 385, 193, 198, 90, 329, 137, 200, 80, 355, 317, 33, 64, 213, 368, 49, 435, 134, 254, 330, 60, 39, 218, 109, 301, 133, 328, 181, 156, 395, 43, 249, 27, 153, 211, 420, 366, 186, 163, 63, 202, 45, 69, 31, 167, 177, 95, 151, 392, 387, 18, 286, 102, 290, 195, 428, 403, 97, 406, 412, 374, 263, 371, 41, 212, 52, 238, 345, 4, 117, 407, 56, 103, 118, 319, 57, 128, 294, 15, 273, 327, 281, 378, 121, 113, 284, 422, 389, 216, 58, 418, 309, 123, 116, 222, 122, 11, 361, 179, 255, 239, 225, 381, 424, 343, 287, 341, 184, 380, 196, 174, 306, 252, 148, 416, 391, 25, 208, 59, 364])
-
-
-
enemy_race(replay)int321 1 2 2 1 1 2 1 ... 2 2 0 2 2 0 0 2
array([1, 1, 2, 2, 1, 1, 2, 1, 1, 2, 2, 2, 1, 1, 0, 2, 2, 0, 0, 2, 0, 2, 2, 1, 1, 2, 2, 0, 0, 0, 0, 0, 0, 2, 0, 1, 2, 1, 1, 2, 2, 2, 1, 0, 0, 1, 0, 2, 2, 2, 1, 0, 2, 1, 0, 2, 1, 0, 1, 1, 2, 2, 2, 2, 1, 2, 1, 2, 0, 0, 0, 1, 0, 2, 2, 0, 0, 0, 0, 1, 0, 1, 2, 0, 1, 2, 2, 1, 2, 1, 0, 2, 2, 2, 1, 0, 0, 2, 1, 1, 1, 1, 0, 0, 0, 1, 0, 2, 0, 2, 0, 1, 0, 0, 2, 0, 0, 0, 2, 1, 0, 1, 1, 0, 2, 1, 0, 2, 0, 1, 2, 2, 0, 2, 2, 0, 0, 2], dtype=int32)
-
enemy_mmr(replay)int323748 3946 3802 ... 3977 3937 4240
array([3748, 3946, 3802, 3816, 3701, 3861, 3907, 4066, 3757, 4350, 3964, 3894, 3932, 3952, 3955, 3912, 3744, 3492, 4642, 3756, 3664, 3769, 3871, 3631, 3631, 3685, 3626, 3647, 3626, 3736, 3784, 3856, 3866, 3847, 3883, 3761, 3686, 3827, 3771, 3752, 3803, 3922, 3888, 3900, 3962, 3992, 3798, 3797, 3790, 3966, 3881, 3823, 3904, 3920, 3907, 3961, 3874, 3570, 4359, 4009, 3859, 3988, 3885, 4020, 3935, 3908, 4596, 4851, 3941, 3808, 3942, 3854, 3956, 3976, 3998, 3953, 4068, 3960, 4016, 4055, 3925, 4104, 3989, 4001, 4237, 4161, 4024, 4108, 4125, 4028, 4012, 3992, 4144, 4029, 4051, 4131, 4407, 4037, 3965, 3944, 4169, 3882, 4018, 4066, 3828, 3957, 4181, 4160, 3886, 3963, 3875, 3956, 3976, 3953, 3757, 4007, 3911, 3967, 4935, 3898, 3880, 3927, 3948, 3833, 3894, 3782, 3786, 3845, 3863, 3966, 3864, 3843, 3840, 4033, 3914, 3977, 3937, 4240], dtype=int32)
-
-
- created_at :
- 2020-06-20T14:25:20.607687
- arviz_version :
- 0.8.3
- inference_library :
- pymc3
- inference_library_version :
- 3.9.1
xarray.Dataset
-
Looks like that helped!
az.plot_trace(output, var_names=var_names);
We can also plot the posteriors and the Highest Density Intervals, whose value for the threshold of 94% says that our model is 94% sure my PvT average MMR is located between 4045 and 4489:
az.plot_posterior(output, var_names = var_names);
This already shows us two things:
- there are large differences in average MMR in the three matchups - as we had expected! This basically means I could go to TopTierPractice and use these MMR ranges to find practice partners at appropriate levels for the both of us. Interestingly, this means I should have been seeking better Terran players. (Note that this is last year's data).
- The versus Zerg fluctuations are a tad larger than against the other races; this would mean there are more confounding variables. Game duration could be a factor; I feel much more confident in the midgame than in the extreme lategame or against early rushes.
And this is honestly something you could start applying to your own data right now. Now, what I was going to do was to also apply a hierarchical model to this data; but, since it turns out I don't really understand them all that well just yet, and I'm running out of time for my pre-set deadling for this post, I'll postpone that for the next week.
Instead, what we'll do is take a look at the current (this year's) data, using this exact same technique for a cheap replica of time dependence:
Analysis repeated for 2020 (easy code snippet!)¶
data = all_data[(all_data['time_played_at'] > '2020-01-01')]
display(data)
time_played_at | win | race | enemy_race | mmr | mmr_diff | enemy_nickname | map_name | duration | enemy_mmr | expected_winrate | |
---|---|---|---|---|---|---|---|---|---|---|---|
265 | 2020-01-04 19:04:15+00:00 | True | Protoss | Terran | 4009 | -29 | Rêgo | Nightshade LE | 935 | 4038 | 0.481039 |
223 | 2020-01-04 19:07:07+00:00 | True | Protoss | Protoss | 4031 | -132 | <unƊeaƊ><sp/>òMégà | World of Sleepers LE | 107 | 4163 | 0.414501 |
402 | 2020-01-04 19:19:31+00:00 | False | Protoss | Terran | 4057 | -32 | Tsumi | Eternal Empire LE | 714 | 4089 | 0.479080 |
206 | 2020-01-05 14:03:19+00:00 | False | Protoss | Zerg | 4036 | -160 | <Lowko><sp/>HardNeper | Eternal Empire LE | 300 | 4196 | 0.396839 |
417 | 2020-01-05 14:18:07+00:00 | True | Protoss | Terran | 4019 | 14 | delmak | Eternal Empire LE | 867 | 4005 | 0.509157 |
... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... | ... |
1 | 2020-06-09 17:11:15+00:00 | False | Protoss | Zerg | 4186 | 39 | djakette | Eternal Empire LE | 420 | 4147 | 0.525489 |
210 | 2020-06-09 17:24:17+00:00 | False | Protoss | Terran | 4164 | -68 | StaMinA | Eternal Empire LE | 758 | 4232 | 0.455635 |
188 | 2020-06-10 11:18:56+00:00 | False | Protoss | Protoss | 4144 | -41 | Guinness | Ice and Chrome LE | 299 | 4185 | 0.473206 |
131 | 2020-06-10 11:27:56+00:00 | False | Protoss | Terran | 4125 | 125 | Oink | Eternal Empire LE | 211 | 4000 | 0.581047 |
144 | 2020-06-10 11:35:16+00:00 | False | Protoss | Protoss | 4100 | -88 | <UATeam><sp/>click | Submarine LE | 343 | 4188 | 0.442688 |
202 rows × 11 columns
brush = altair.selection(type='interval')
scatter = altair.Chart(data).mark_circle().encode(
altair.X('enemy_mmr',
scale=altair.Scale(zero=False)),
altair.Y('mmr',
scale=altair.Scale(zero=False)),
facet='enemy_race',
size='expected_winrate',
color='win',
tooltip='enemy_nickname',
).add_selection(brush)
bar = altair.Chart(data).mark_bar().encode(
x=altair.X('mean(win):Q', scale=altair.Scale(domain=(0, 1))),
y='enemy_race:O',
).transform_filter(brush)
scatter & bar
import pymc3 as pm
import arviz as az
coords = {
"replay": data.index,
"race": ["Terran", "Protoss", "Zerg"],
}
race_encoding ={"Terran": 0,
"Protoss": 1,
"Zerg": 2}
predictive_var_names = "win μ σ winrate".split()
with pm.Model(coords=coords) as split_model:
mmr_μ_matchup = pm.Normal('μ', 4000, 300, dims='race')
mmr_σ_matchup = pm.HalfNormal('σ', 100, dims='race')
mmr_σ_norm = pm.Normal('helper', 0, 1, dims='replay')
enemy_races = pm.Data("enemy_race", data.enemy_race.map(race_encoding).astype(int), dims='replay')
mmr = pm.Deterministic('MMR', mmr_μ_matchup[enemy_races] + mmr_σ_matchup[enemy_races] * mmr_σ_norm, dims='replay')
enemy_mmr = pm.Data("enemy_mmr", data.enemy_mmr, dims='replay')
diffs = pm.Deterministic('MMR_diff', mmr - enemy_mmr, dims = 'replay')
p = pm.Deterministic('winrate', MMR_winrate(diffs), dims = 'replay')
wl = pm.Bernoulli('win', p=p, observed=data.win, dims = 'replay')
trace = pm.sample(2000, tune=2000, chains=4, random_seed=1, target_accept=0.9)
output = az.from_pymc3(trace=trace,
prior=pm.sample_prior_predictive(2000 , var_names=predictive_var_names, random_seed=1),
posterior_predictive=pm.sample_posterior_predictive(trace, var_names=predictive_var_names, random_seed=1),
)
output
Auto-assigning NUTS sampler... Initializing NUTS using jitter+adapt_diag... Multiprocess sampling (4 chains in 2 jobs) NUTS: [helper, σ, μ]
Sampling 4 chains for 2_000 tune and 2_000 draw iterations (8_000 + 8_000 draws total) took 26 seconds.
-
-
- chain: 4
- draw: 2000
- race: 3
- replay: 202
-
-
chain(chain)int640 1 2 3
array([0, 1, 2, 3])
-
draw(draw)int640 1 2 3 4 ... 1996 1997 1998 1999
array([ 0, 1, 2, ..., 1997, 1998, 1999])
-
race(race)<U7'Terran' 'Protoss' 'Zerg'
array(['Terran', 'Protoss', 'Zerg'], dtype='<U7')
-
replay(replay)int64265 223 402 206 ... 210 188 131 144
array([265, 223, 402, ..., 188, 131, 144])
-
-
-
μ(chain, draw, race)float644.211e+03 4.108e+03 ... 3.992e+03
array([[[4210.85745791, 4108.13402572, 4047.68655821], [4187.82039211, 4167.02974265, 3832.39048778], [4385.10730355, 4051.15698963, 3888.3543684 ], ..., [4268.67850837, 4004.06007755, 3935.24654733], [4182.61327252, 3852.96115761, 4013.25955458], [4058.95730755, 3860.61071343, 4048.4206048 ]], [[4331.22003251, 3887.80946609, 3896.23173387], [4106.02686452, 4063.88927972, 3995.75729977], [4196.19055581, 3907.28257413, 3817.30936413], ..., [4158.82110597, 3994.77797297, 3948.94402472], [4130.32958212, 4024.45281323, 3864.17620077], [4255.71656811, 4121.39660228, 4067.50723152]], [[4264.33592228, 4119.64071496, 3900.73678858], [4290.27572226, 4138.58898858, 3923.24665912], [4057.54009725, 3929.51218737, 3995.78802404], ..., [4271.91923915, 4208.66941361, 3925.91501886], [4170.24764659, 4004.86306088, 3916.82405953], [4146.33857085, 4069.27579897, 3925.21194574]], [[4391.8074459 , 3973.03474104, 4062.09104669], [3980.12121482, 4056.28919035, 3860.08003276], [4365.43547727, 4060.16002941, 3927.51863056], ..., [4344.05048768, 3950.01257309, 3857.26550133], [3986.12083546, 4073.15663108, 4093.19655361], [4133.45346107, 4052.34439539, 3991.74262184]]])
-
helper(chain, draw, replay)float640.8451 -0.2949 ... -0.3081 0.4416
array([[[ 0.84509895, -0.29485376, 0.8011897 , ..., 1.33714031, -1.80845999, -0.51358847], [-1.40481271, 1.07728824, -1.1490509 , ..., -0.41919933, 1.34160767, -0.04753472], [-1.44831415, -0.00634847, -0.67124799, ..., -1.08190966, -0.64604433, -0.69881696], ..., [-1.01416874, 1.77692415, -0.0175506 , ..., 0.02982226, -0.79090697, -0.75716479], [ 0.99411964, -0.39523617, 0.46654021, ..., -1.62395586, -0.63782079, 1.43213844], [ 1.77132541, -0.95056043, 0.37540258, ..., 0.01968714, -1.03527548, 0.39205272]], [[-0.10334535, -0.8929405 , 0.73607739, ..., -1.19666118, 0.06852772, -0.96987602], [ 2.13759694, 0.84869979, -0.99459824, ..., 0.65078661, -0.49690394, 0.77831748], [-1.44472736, -1.21900212, -0.77229333, ..., -0.45423455, 0.19830191, -0.68767208], ..., [ 0.19454783, -0.71190753, 2.82909697, ..., 0.97857095, 0.67187882, 0.09746846], [-0.58904084, -0.41505728, 1.04412233, ..., -0.12198221, -0.50684467, 1.56542232], [-0.89195114, 1.23244903, -2.10987152, ..., 0.91622949, 0.40999001, -0.65424188]], [[ 0.95626333, 0.21940413, 0.86589926, ..., 1.47272558, -0.55443001, 1.2008141 ], [ 0.94032427, 0.0931288 , 0.77226228, ..., 1.26916956, -0.57662136, 0.88842902], [-0.23138386, 0.15845085, -1.33668801, ..., -1.33332178, 0.15524309, -1.08860061], ..., [ 1.26120065, -0.15534739, 0.34744789, ..., -1.53003928, -0.07152691, -1.72074586], [-1.15880002, 0.73396249, -0.10402144, ..., 1.01286394, -0.78903161, 2.12732981], [-0.12447582, -0.07029113, -0.5904053 , ..., 0.45058805, 0.23860171, 1.39694982]], [[ 1.37288017, 1.57454868, -1.31150848, ..., -0.11459006, 1.57652361, 0.32320881], [-1.15566697, -0.90959159, 1.06482753, ..., -0.56897759, -1.87990178, -0.52078417], [-0.1134551 , -0.66397135, -0.29449925, ..., 1.63000314, 0.50064175, -0.64631669], ..., [ 0.57673475, 1.49125537, 0.97739831, ..., -0.26974367, -1.55536186, -0.15896637], [-0.18430432, 0.11944914, -0.56572547, ..., -0.51577881, -0.14535604, -1.09840927], [ 0.76741937, 1.39186527, 1.48469105, ..., 0.19037821, -0.30807854, 0.44160543]]])
-
σ(chain, draw, race)float6427.23 6.505 39.42 ... 47.14 64.05
array([[[ 27.22582328, 6.50453839, 39.42478265], [ 13.37751425, 8.46802428, 129.89610941], [ 10.72164228, 24.94971223, 102.7146095 ], ..., [107.26940159, 23.20508653, 35.3446602 ], [124.97180552, 13.19227419, 145.18684899], [ 88.2747096 , 40.61048868, 50.21903074]], [[131.31729221, 96.5412831 , 70.56487648], [143.55608877, 36.54926803, 5.84418576], [ 33.35705402, 205.64975605, 10.01984894], ..., [ 70.90358774, 23.20074938, 104.54547733], [ 74.54938645, 48.98896964, 131.8366626 ], [ 28.87932693, 103.09557275, 45.74619814]], [[ 98.66459786, 213.93501451, 77.75342308], [ 46.20775227, 140.47053288, 87.54119877], [ 48.65507021, 182.63370476, 35.63529879], ..., [ 4.05083801, 116.69219995, 127.42758537], [ 28.45085473, 79.66533597, 140.58625841], [148.66125575, 154.45834609, 239.20924515]], [[ 22.20125272, 74.76545816, 94.2220565 ], [ 63.23463911, 29.49654959, 78.6400277 ], [ 94.44434706, 67.93754801, 187.33159977], ..., [ 94.74513333, 55.86780229, 124.72461746], [ 14.08098885, 94.3565778 , 120.11072466], [ 18.22473629, 47.14089139, 64.04793555]]])
-
MMR(chain, draw, replay)float644.234e+03 4.106e+03 ... 4.073e+03
array([[[4233.86597263, 4106.21613813, 4232.67050709, ..., 4116.83150619, 4161.62064575, 4104.79336979], [4169.02748999, 4176.15224566, 4172.44894728, ..., 4163.47995259, 4205.76776789, 4166.62721748], [4369.57899735, 4050.99859722, 4377.91042278, ..., 4024.16365495, 4378.18064731, 4033.72170767], ..., [4159.88923469, 4045.29375632, 4266.79586648, ..., 4004.75210576, 4183.8383908 , 3986.49000311], [4306.85019842, 3847.74709363, 4240.91764535, ..., 3831.53748658, 4102.90365662, 3871.85432064], [4215.32054396, 3822.00798974, 4092.09586122, ..., 3861.41021769, 3967.5686653 , 3876.53216602]], [[4317.64900043, 3801.60384413, 4427.87972211, ..., 3772.28226007, 4340.21890649, 3794.17639083], [4412.89192094, 4094.90863597, 3963.24623187, ..., 4087.6750538 , 4034.6932786 , 4092.33621384], [4147.99870736, 3656.59508486, 4170.42912556, ..., 3813.86934982, 4202.80532332, 3765.86297953], ..., [4172.61524529, 3978.2611847 , 4359.41423107, ..., 4017.48155239, 4206.45972478, 3997.0393142 ], [4086.41694857, 4004.11958463, 4208.1682613 , ..., 4018.47703031, 4092.54462294, 4101.14123968], [4229.95761948, 4248.45664107, 4194.78489881, ..., 4215.85580654, 4267.5568035 , 4053.94716107]], [[4358.6852591 , 4166.57894108, 4349.76952448, ..., 4434.70828399, 4209.63330835, 4376.53689592], [4333.72599335, 4151.67084096, 4325.96022654, ..., 4316.86991295, 4263.63134537, 4263.38708701], [4046.28209913, 3958.4506536 , 3992.50344832, ..., 3686.00269018, 4065.09346086, 3730.69702582], ..., [4277.0281587 , 4190.54158446, 4273.32669429, ..., 4030.12576439, 4271.62949522, 4007.87179371], [4137.27879557, 4063.33442943, 4167.28814765, ..., 4085.55320713, 4147.79902283, 4174.33750488], [4127.8338395 , 4058.4187474 , 4058.56817809, ..., 4138.87288461, 4181.8094006 , 4285.04635831]], [[4422.28710553, 4090.75659466, 4362.69031474, ..., 3964.46736236, 4426.80824501, 3997.1995961 ], [3907.04303111, 4029.45937695, 4047.45519966, ..., 4039.50631474, 3861.24630434, 4040.92785429], [4354.72028456, 4015.05144374, 4337.62168777, ..., 4170.8984463 , 4412.71826075, 4016.25085813], ..., [4398.69329839, 4033.32573322, 4436.65422109, ..., 3934.9425872 , 4196.68752121, 3941.13147145], [3983.52564834, 4084.42744336, 3978.15486147, ..., 4024.48950801, 3984.07407872, 3969.51449166], [4147.43947677, 4117.95816506, 4160.51156385, ..., 4061.31899411, 4127.83881087, 4073.1620688 ]]])
-
MMR_diff(chain, draw, replay)float64195.9 -56.78 143.7 ... 127.8 -114.8
array([[[ 195.86597263, -56.78386187, 143.67050709, ..., -68.16849381, 161.62064575, -83.20663021], [ 131.02748999, 13.15224566, 83.44894728, ..., -21.52004741, 205.76776789, -21.37278252], [ 331.57899735, -112.00140278, 288.91042278, ..., -160.83634505, 378.18064731, -154.27829233], ..., [ 121.88923469, -117.70624368, 177.79586648, ..., -180.24789424, 183.8383908 , -201.50999689], [ 268.85019842, -315.25290637, 151.91764535, ..., -353.46251342, 102.90365662, -316.14567936], [ 177.32054396, -340.99201026, 3.09586122, ..., -323.58978231, -32.4313347 , -311.46783398]], [[ 279.64900043, -361.39615587, 338.87972211, ..., -412.71773993, 340.21890649, -393.82360917], [ 374.89192094, -68.09136403, -125.75376813, ..., -97.3249462 , 34.6932786 , -95.66378616], [ 109.99870736, -506.40491514, 81.42912556, ..., -371.13065018, 202.80532332, -422.13702047], ..., [ 134.61524529, -184.7388153 , 270.41423107, ..., -167.51844761, 206.45972478, -190.9606858 ], [ 48.41694857, -158.88041537, 119.1682613 , ..., -166.52296969, 92.54462294, -86.85876032], [ 191.95761948, 85.45664107, 105.78489881, ..., 30.85580654, 267.5568035 , -134.05283893]], [[ 320.6852591 , 3.57894108, 260.76952448, ..., 249.70828399, 209.63330835, 188.53689592], [ 295.72599335, -11.32915904, 236.96022654, ..., 131.86991295, 263.63134537, 75.38708701], [ 8.28209913, -204.5493464 , -96.49655168, ..., -498.99730982, 65.09346086, -457.30297418], ..., [ 239.0281587 , 27.54158446, 184.32669429, ..., -154.87423561, 271.62949522, -180.12820629], [ 99.27879557, -99.66557057, 78.28814765, ..., -99.44679287, 147.79902283, -13.66249512], [ 89.8338395 , -104.5812526 , -30.43182191, ..., -46.12711539, 181.8094006 , 97.04635831]], [[ 384.28710553, -72.24340534, 273.69031474, ..., -220.53263764, 426.80824501, -190.8004039 ], [-130.95696889, -133.54062305, -41.54480034, ..., -145.49368526, -138.75369566, -147.07214571], [ 316.72028456, -147.94855626, 248.62168777, ..., -14.1015537 , 412.71826075, -171.74914187], ..., [ 360.69329839, -129.67426678, 347.65422109, ..., -250.0574128 , 196.68752121, -246.86852855], [ -54.47435166, -78.57255664, -110.84513853, ..., -160.51049199, -15.92592128, -218.48550834], [ 109.43947677, -45.04183494, 71.51156385, ..., -123.68100589, 127.83881087, -114.8379312 ]]])
-
winrate(chain, draw, replay)float640.6254 0.4629 ... 0.5829 0.4254
array([[[0.62539184, 0.46292339, 0.59288977, ..., 0.45552587, 0.60417516, 0.44578491], [0.58488098, 0.50860261, 0.55437173, ..., 0.48592652, 0.63144167, 0.48602278], [0.70424627, 0.4272549 , 0.68047498, ..., 0.39631572, 0.72899659, 0.40042837], ..., [0.579064 , 0.42360614, 0.61425083, ..., 0.38423006, 0.6179903 , 0.37115482], [0.66895588, 0.30472821, 0.59808776, ..., 0.28396851, 0.56691001, 0.30423351], [0.6139561 , 0.29064968, 0.50202513, ..., 0.30012628, 0.47879797, 0.3068306 ]], [[0.6751831 , 0.27976657, 0.7082095 , ..., 0.25352371, 0.70893308, 0.26299327], [0.72729319, 0.45557593, 0.41847324, ..., 0.43667736, 0.52267881, 0.43774686], [0.57146229, 0.20998001, 0.55306573, ..., 0.27466314, 0.6296359 , 0.24888782], ..., [0.58715842, 0.38145365, 0.66986153, ..., 0.3921403 , 0.63186293, 0.37761992], [0.53162934, 0.3975408 , 0.57732763, ..., 0.39276136, 0.56024338, 0.44342522], [0.62299295, 0.55566915, 0.56876006, ..., 0.52017317, 0.66820599, 0.41319834]], [[0.69827494, 0.50234112, 0.66425691, ..., 0.65777178, 0.6337924 , 0.6208884 ], [0.68433995, 0.49258965, 0.650224 , ..., 0.58541607, 0.66592486, 0.54915469], [0.50541747, 0.36930058, 0.43721063, ..., 0.21321342, 0.54247782, 0.23208648], ..., [0.65145361, 0.51800836, 0.61829188, ..., 0.40005405, 0.67056436, 0.38430415], [0.56457983, 0.4351714 , 0.55103335, ..., 0.43531211, 0.59549456, 0.49106372], [0.55849514, 0.43201257, 0.48010373, ..., 0.46986282, 0.61673617, 0.56314332]], [[0.73214164, 0.45288266, 0.67175446, ..., 0.35961398, 0.75339009, 0.3777185 ], [0.41516382, 0.41352334, 0.47285047, ..., 0.40595928, 0.4102192 , 0.40496366], [0.69608465, 0.40441119, 0.65713147, ..., 0.49077661, 0.74647655, 0.38950478], ..., [0.71986271, 0.41597896, 0.71293119, ..., 0.34202261, 0.62589531, 0.34390283], [0.46442617, 0.44878255, 0.42799542, ..., 0.39651972, 0.48958367, 0.36084846], [0.57110391, 0.47057023, 0.54664281, ..., 0.41979366, 0.58285382, 0.42543967]]])
-
-
- created_at :
- 2020-06-20T14:41:39.607220
- arviz_version :
- 0.8.3
- inference_library :
- pymc3
- inference_library_version :
- 3.9.1
- sampling_time :
- 25.67070198059082
- tuning_steps :
- 2000
xarray.Dataset
-
-
-
- chain: 4
- draw: 2000
- race: 3
- replay: 202
-
-
chain(chain)int640 1 2 3
array([0, 1, 2, 3])
-
draw(draw)int640 1 2 3 4 ... 1996 1997 1998 1999
array([ 0, 1, 2, ..., 1997, 1998, 1999])
-
replay(replay)int64265 223 402 206 ... 210 188 131 144
array([265, 223, 402, ..., 188, 131, 144])
-
race(race)<U7'Terran' 'Protoss' 'Zerg'
array(['Terran', 'Protoss', 'Zerg'], dtype='<U7')
-
-
-
win(chain, draw, replay)int641 1 1 0 1 1 1 1 ... 0 0 0 0 1 0 1 0
array([[[1, 1, 1, ..., 1, 0, 1], [0, 0, 1, ..., 1, 1, 1], [1, 1, 1, ..., 0, 1, 1], ..., [1, 0, 1, ..., 0, 0, 0], [1, 0, 0, ..., 1, 0, 0], [1, 0, 0, ..., 1, 0, 0]], [[0, 1, 0, ..., 0, 0, 0], [1, 0, 0, ..., 0, 0, 0], [0, 0, 0, ..., 0, 1, 0], ..., [1, 1, 0, ..., 1, 1, 1], [1, 0, 1, ..., 0, 1, 0], [1, 1, 1, ..., 1, 1, 1]], [[1, 0, 1, ..., 1, 1, 0], [0, 0, 1, ..., 0, 1, 1], [1, 0, 1, ..., 0, 1, 0], ..., [0, 1, 1, ..., 1, 0, 1], [1, 0, 1, ..., 0, 0, 1], [0, 0, 1, ..., 1, 1, 0]], [[1, 1, 0, ..., 0, 1, 0], [0, 0, 0, ..., 0, 1, 0], [1, 1, 1, ..., 1, 1, 0], ..., [1, 0, 1, ..., 0, 1, 0], [0, 0, 0, ..., 1, 1, 1], [1, 1, 1, ..., 0, 1, 0]]])
-
μ(chain, draw, race)float644.211e+03 4.108e+03 ... 3.992e+03
array([[[4210.85745791, 4108.13402572, 4047.68655821], [4187.82039211, 4167.02974265, 3832.39048778], [4385.10730355, 4051.15698963, 3888.3543684 ], ..., [4268.67850837, 4004.06007755, 3935.24654733], [4182.61327252, 3852.96115761, 4013.25955458], [4058.95730755, 3860.61071343, 4048.4206048 ]], [[4331.22003251, 3887.80946609, 3896.23173387], [4106.02686452, 4063.88927972, 3995.75729977], [4196.19055581, 3907.28257413, 3817.30936413], ..., [4158.82110597, 3994.77797297, 3948.94402472], [4130.32958212, 4024.45281323, 3864.17620077], [4255.71656811, 4121.39660228, 4067.50723152]], [[4264.33592228, 4119.64071496, 3900.73678858], [4290.27572226, 4138.58898858, 3923.24665912], [4057.54009725, 3929.51218737, 3995.78802404], ..., [4271.91923915, 4208.66941361, 3925.91501886], [4170.24764659, 4004.86306088, 3916.82405953], [4146.33857085, 4069.27579897, 3925.21194574]], [[4391.8074459 , 3973.03474104, 4062.09104669], [3980.12121482, 4056.28919035, 3860.08003276], [4365.43547727, 4060.16002941, 3927.51863056], ..., [4344.05048768, 3950.01257309, 3857.26550133], [3986.12083546, 4073.15663108, 4093.19655361], [4133.45346107, 4052.34439539, 3991.74262184]]])
-
σ(chain, draw, race)float6427.23 6.505 39.42 ... 47.14 64.05
array([[[ 27.22582328, 6.50453839, 39.42478265], [ 13.37751425, 8.46802428, 129.89610941], [ 10.72164228, 24.94971223, 102.7146095 ], ..., [107.26940159, 23.20508653, 35.3446602 ], [124.97180552, 13.19227419, 145.18684899], [ 88.2747096 , 40.61048868, 50.21903074]], [[131.31729221, 96.5412831 , 70.56487648], [143.55608877, 36.54926803, 5.84418576], [ 33.35705402, 205.64975605, 10.01984894], ..., [ 70.90358774, 23.20074938, 104.54547733], [ 74.54938645, 48.98896964, 131.8366626 ], [ 28.87932693, 103.09557275, 45.74619814]], [[ 98.66459786, 213.93501451, 77.75342308], [ 46.20775227, 140.47053288, 87.54119877], [ 48.65507021, 182.63370476, 35.63529879], ..., [ 4.05083801, 116.69219995, 127.42758537], [ 28.45085473, 79.66533597, 140.58625841], [148.66125575, 154.45834609, 239.20924515]], [[ 22.20125272, 74.76545816, 94.2220565 ], [ 63.23463911, 29.49654959, 78.6400277 ], [ 94.44434706, 67.93754801, 187.33159977], ..., [ 94.74513333, 55.86780229, 124.72461746], [ 14.08098885, 94.3565778 , 120.11072466], [ 18.22473629, 47.14089139, 64.04793555]]])
-
winrate(chain, draw, replay)float640.6254 0.4629 ... 0.5829 0.4254
array([[[0.62539184, 0.46292339, 0.59288977, ..., 0.45552587, 0.60417516, 0.44578491], [0.58488098, 0.50860261, 0.55437173, ..., 0.48592652, 0.63144167, 0.48602278], [0.70424627, 0.4272549 , 0.68047498, ..., 0.39631572, 0.72899659, 0.40042837], ..., [0.579064 , 0.42360614, 0.61425083, ..., 0.38423006, 0.6179903 , 0.37115482], [0.66895588, 0.30472821, 0.59808776, ..., 0.28396851, 0.56691001, 0.30423351], [0.6139561 , 0.29064968, 0.50202513, ..., 0.30012628, 0.47879797, 0.3068306 ]], [[0.6751831 , 0.27976657, 0.7082095 , ..., 0.25352371, 0.70893308, 0.26299327], [0.72729319, 0.45557593, 0.41847324, ..., 0.43667736, 0.52267881, 0.43774686], [0.57146229, 0.20998001, 0.55306573, ..., 0.27466314, 0.6296359 , 0.24888782], ..., [0.58715842, 0.38145365, 0.66986153, ..., 0.3921403 , 0.63186293, 0.37761992], [0.53162934, 0.3975408 , 0.57732763, ..., 0.39276136, 0.56024338, 0.44342522], [0.62299295, 0.55566915, 0.56876006, ..., 0.52017317, 0.66820599, 0.41319834]], [[0.69827494, 0.50234112, 0.66425691, ..., 0.65777178, 0.6337924 , 0.6208884 ], [0.68433995, 0.49258965, 0.650224 , ..., 0.58541607, 0.66592486, 0.54915469], [0.50541747, 0.36930058, 0.43721063, ..., 0.21321342, 0.54247782, 0.23208648], ..., [0.65145361, 0.51800836, 0.61829188, ..., 0.40005405, 0.67056436, 0.38430415], [0.56457983, 0.4351714 , 0.55103335, ..., 0.43531211, 0.59549456, 0.49106372], [0.55849514, 0.43201257, 0.48010373, ..., 0.46986282, 0.61673617, 0.56314332]], [[0.73214164, 0.45288266, 0.67175446, ..., 0.35961398, 0.75339009, 0.3777185 ], [0.41516382, 0.41352334, 0.47285047, ..., 0.40595928, 0.4102192 , 0.40496366], [0.69608465, 0.40441119, 0.65713147, ..., 0.49077661, 0.74647655, 0.38950478], ..., [0.71986271, 0.41597896, 0.71293119, ..., 0.34202261, 0.62589531, 0.34390283], [0.46442617, 0.44878255, 0.42799542, ..., 0.39651972, 0.48958367, 0.36084846], [0.57110391, 0.47057023, 0.54664281, ..., 0.41979366, 0.58285382, 0.42543967]]])
-
-
- created_at :
- 2020-06-20T14:41:40.227565
- arviz_version :
- 0.8.3
- inference_library :
- pymc3
- inference_library_version :
- 3.9.1
xarray.Dataset
-
-
-
- chain: 4
- draw: 2000
- replay: 202
-
-
chain(chain)int640 1 2 3
array([0, 1, 2, 3])
-
draw(draw)int640 1 2 3 4 ... 1996 1997 1998 1999
array([ 0, 1, 2, ..., 1997, 1998, 1999])
-
replay(replay)int64265 223 402 206 ... 210 188 131 144
array([265, 223, 402, ..., 188, 131, 144])
-
-
-
win(chain, draw, replay)float64-0.4694 -0.7702 ... -0.8743 -0.5542
array([[[-0.46937689, -0.7701937 , -0.89867129, ..., -0.60793486, -0.92678349, -0.59020241], [-0.5363469 , -0.6760883 , -0.80827016, ..., -0.66538907, -0.99815629, -0.66557633], [-0.35062718, -0.85037448, -1.14091971, ..., -0.50470393, -1.30562387, -0.51153982], ..., [-0.54634227, -0.85895116, -0.95256795, ..., -0.48488185, -0.96230927, -0.46387019], [-0.40203718, -1.188335 , -0.91152153, ..., -0.33403113, -0.83680974, -0.36274118], [-0.48783186, -1.23563659, -0.69720566, ..., -0.35685536, -0.65161755, -0.36648086]], [[-0.39277136, -1.27379969, -1.23171919, ..., -0.29239143, -1.23420208, -0.30515825], [-0.31842559, -0.78619288, -0.54209829, ..., -0.57390274, -0.73956566, -0.57580311], [-0.55955678, -1.56074295, -0.80534375, ..., -0.32111909, -0.9932687 , -0.28620026], ..., [-0.53246061, -0.96376593, -1.1082431 , ..., -0.49781118, -0.99929993, -0.47420432], [-0.63180876, -0.9224577 , -0.86115793, ..., -0.49883342, -0.82153384, -0.58595374], [-0.47322008, -0.58758221, -0.84109064, ..., -0.73433001, -1.10324095, -0.5330684 ]], [[-0.35914236, -0.68847586, -1.09140904, ..., -1.07227745, -1.00455489, -0.96992466], [-0.37930049, -0.70807881, -1.05046232, ..., -0.88047984, -1.09638935, -0.796631 ], [-0.68237052, -0.99614438, -0.57484984, ..., -0.23979825, -0.78192993, -0.26407816], ..., [-0.42854909, -0.65776391, -0.96309906, ..., -0.51091572, -1.11037426, -0.48500219], [-0.57167349, -0.8320153 , -0.80080667, ..., -0.57148211, -0.9050901 , -0.67543246], [-0.58250936, -0.83930058, -0.65412597, ..., -0.63461948, -0.95903169, -0.8281501 ]], [[-0.31178129, -0.79212221, -1.11399335, ..., -0.44568413, -1.39994749, -0.47436271], [-0.8790821 , -0.88304132, -0.64027103, ..., -0.52080742, -0.52800433, -0.5191328 ], [-0.36228401, -0.90532311, -1.07040821, ..., -0.67486847, -1.37229894, -0.49348481], ..., [-0.32869477, -0.87712059, -1.24803335, ..., -0.41858471, -0.9832196 , -0.42144637], [-0.76695269, -0.80121681, -0.55860828, ..., -0.50504192, -0.67252855, -0.4476137 ], [-0.56018411, -0.75381006, -0.79107498, ..., -0.54437148, -0.87431857, -0.55415018]]])
-
-
- created_at :
- 2020-06-20T14:41:40.224319
- arviz_version :
- 0.8.3
- inference_library :
- pymc3
- inference_library_version :
- 3.9.1
xarray.Dataset
-
-
-
- chain: 4
- draw: 2000
-
-
chain(chain)int640 1 2 3
array([0, 1, 2, 3])
-
draw(draw)int640 1 2 3 4 ... 1996 1997 1998 1999
array([ 0, 1, 2, ..., 1997, 1998, 1999])
-
-
-
step_size_bar(chain, draw)float640.2948 0.2948 ... 0.2904 0.2904
array([[0.29475372, 0.29475372, 0.29475372, ..., 0.29475372, 0.29475372, 0.29475372], [0.29770566, 0.29770566, 0.29770566, ..., 0.29770566, 0.29770566, 0.29770566], [0.27693519, 0.27693519, 0.27693519, ..., 0.27693519, 0.27693519, 0.27693519], [0.29040166, 0.29040166, 0.29040166, ..., 0.29040166, 0.29040166, 0.29040166]])
-
energy_error(chain, draw)float64-0.2597 0.03461 ... 0.2249 0.4218
array([[-0.25966 , 0.03460919, 0.01227495, ..., -0.51621275, 0.86212363, -0.24459744], [-0.22234907, 0.28075065, -0.17805517, ..., 0.24543193, -0.28516645, 0.2238124 ], [-1.07546919, 0.10968526, 0.01985628, ..., -0.48126788, -0.01111381, -0.44255399], [ 0.0773235 , 0.06891559, -0.09303872, ..., 0.02184957, 0.2248607 , 0.42175383]])
-
step_size(chain, draw)float640.3427 0.3427 ... 0.3262 0.3262
array([[0.34270119, 0.34270119, 0.34270119, ..., 0.34270119, 0.34270119, 0.34270119], [0.24183475, 0.24183475, 0.24183475, ..., 0.24183475, 0.24183475, 0.24183475], [0.30871193, 0.30871193, 0.30871193, ..., 0.30871193, 0.30871193, 0.30871193], [0.32619857, 0.32619857, 0.32619857, ..., 0.32619857, 0.32619857, 0.32619857]])
-
max_energy_error(chain, draw)float640.6081 0.6931 ... 0.2691 0.7324
array([[ 0.60811894, 0.6931234 , -0.45286992, ..., 0.95980239, 1.16150203, -0.51598675], [-0.23944338, 0.28075065, 1.04635515, ..., 0.38439969, -0.57428163, 0.29192163], [-1.07546919, 0.9341971 , 0.44889643, ..., -0.62753125, 0.34544639, -0.49631647], [ 0.14807623, 0.52570743, -0.12783034, ..., 0.52722686, 0.2690612 , 0.73236215]])
-
diverging(chain, draw)boolFalse False False ... False False
array([[False, False, False, ..., False, False, False], [False, False, False, ..., False, False, False], [False, False, False, ..., False, False, False], [False, False, False, ..., False, False, False]])
-
depth(chain, draw)int644 4 4 4 4 4 4 4 ... 4 4 4 4 4 4 4 4
array([[4, 4, 4, ..., 4, 4, 4], [4, 4, 4, ..., 4, 4, 4], [4, 4, 4, ..., 4, 4, 4], [4, 4, 4, ..., 4, 4, 4]])
-
mean_tree_accept(chain, draw)float640.9568 0.9454 ... 0.8909 0.8291
array([[0.95677878, 0.94541275, 0.98043667, ..., 0.95633415, 0.76809655, 0.99642304], [0.99644285, 0.94807192, 0.90218025, ..., 0.82851872, 0.96650925, 0.92233715], [1. , 0.71300232, 0.88258827, ..., 1. , 0.93890528, 1. ], [0.94858298, 0.788294 , 0.98735475, ..., 0.78198705, 0.89093101, 0.8291273 ]])
-
energy(chain, draw)float64571.0 550.5 551.2 ... 535.7 552.8
array([[570.97160285, 550.45584786, 551.24512846, ..., 556.01462471, 539.20361935, 542.4138038 ], [527.22012699, 525.94887852, 554.05126794, ..., 577.34190697, 566.73680041, 554.7874333 ], [551.2205363 , 538.40491567, 540.25295614, ..., 567.3886664 , 554.96612302, 527.45559388], [539.74878505, 563.86859884, 548.89266634, ..., 549.10831899, 535.70152588, 552.7940487 ]])
-
lp(chain, draw)float64-454.4 -444.6 ... -441.8 -450.3
array([[-454.37819213, -444.56013535, -452.36990853, ..., -440.74435605, -445.0448313 , -439.64688243], [-429.31151241, -445.44878934, -454.31044625, ..., -462.85580719, -447.20869749, -457.61499909], [-427.08968302, -433.51041027, -430.05582152, ..., -448.69884569, -444.46669223, -426.50110975], [-441.96784552, -446.14529167, -443.64588735, ..., -436.04827228, -441.77413958, -450.32373305]])
-
tree_size(chain, draw)float6415.0 15.0 15.0 ... 15.0 15.0 15.0
array([[15., 15., 15., ..., 15., 15., 15.], [15., 15., 15., ..., 15., 15., 15.], [15., 15., 15., ..., 15., 15., 15.], [15., 15., 15., ..., 15., 15., 15.]])
-
-
- created_at :
- 2020-06-20T14:41:39.615645
- arviz_version :
- 0.8.3
- inference_library :
- pymc3
- inference_library_version :
- 3.9.1
- sampling_time :
- 25.67070198059082
- tuning_steps :
- 2000
xarray.Dataset
-
-
-
- chain: 1
- draw: 2000
- race: 3
- replay: 202
-
-
chain(chain)int640
array([0])
-
draw(draw)int640 1 2 3 4 ... 1996 1997 1998 1999
array([ 0, 1, 2, ..., 1997, 1998, 1999])
-
race(race)<U7'Terran' 'Protoss' 'Zerg'
array(['Terran', 'Protoss', 'Zerg'], dtype='<U7')
-
replay(replay)int64265 223 402 206 ... 210 188 131 144
array([265, 223, 402, ..., 188, 131, 144])
-
-
-
μ(chain, draw, race)float644.487e+03 3.816e+03 ... 4.079e+03
array([[[4487.3036091 , 3816.4730759 , 3841.54847432], [3678.10941335, 4259.6222888 , 3309.53839094], [4523.44352926, 3771.63792973, 4095.71172882], ..., [3793.7352304 , 3936.80991215, 3897.05804332], [4287.27113474, 4067.38137671, 3856.85673442], [3792.63645476, 3763.38063103, 4079.25539045]]])
-
σ(chain, draw, race)float64203.6 143.1 121.8 ... 137.4 2.277
array([[[203.63022425, 143.11882721, 121.84862917], [ 45.92168871, 54.50152003, 165.24930064], [ 73.03480705, 79.10581705, 32.67507679], ..., [ 2.51873565, 187.68065267, 164.95280152], [128.66435759, 34.45577797, 153.07274428], [120.71958961, 137.40100064, 2.27693063]]])
-
winrate(chain, draw, replay)float640.415 0.3547 ... 0.2377 0.3302
array([[[0.41497111, 0.35469398, 0.71008641, ..., 0.2409493 , 0.6721141 , 0.32116004], [0.29598745, 0.56161731, 0.26909548, ..., 0.55386956, 0.28532212, 0.56595945], [0.82468725, 0.30495278, 0.76779695, ..., 0.28886179, 0.79610061, 0.31071488], ..., [0.34586104, 0.51741056, 0.31504336, ..., 0.31216598, 0.36711597, 0.13365978], [0.69922031, 0.44950387, 0.67782839, ..., 0.41620837, 0.84102901, 0.41210879], [0.35284029, 0.20802614, 0.30976338, ..., 0.35838849, 0.23765418, 0.33020678]]])
-
-
- created_at :
- 2020-06-20T14:41:40.231024
- arviz_version :
- 0.8.3
- inference_library :
- pymc3
- inference_library_version :
- 3.9.1
xarray.Dataset
-
-
-
- chain: 1
- draw: 2000
- replay: 202
-
-
chain(chain)int640
array([0])
-
draw(draw)int640 1 2 3 4 ... 1996 1997 1998 1999
array([ 0, 1, 2, ..., 1997, 1998, 1999])
-
replay(replay)int64265 223 402 206 ... 210 188 131 144
array([265, 223, 402, ..., 188, 131, 144])
-
-
-
win(chain, draw, replay)int640 1 1 0 1 0 1 0 ... 0 0 0 0 0 0 0 0
array([[[0, 1, 1, ..., 0, 1, 0], [0, 1, 0, ..., 0, 1, 0], [1, 0, 0, ..., 1, 1, 0], ..., [0, 1, 0, ..., 1, 1, 0], [1, 0, 1, ..., 1, 0, 1], [1, 0, 0, ..., 0, 0, 0]]])
-
-
- created_at :
- 2020-06-20T14:41:40.232791
- arviz_version :
- 0.8.3
- inference_library :
- pymc3
- inference_library_version :
- 3.9.1
xarray.Dataset
-
-
-
- replay: 202
-
-
replay(replay)int64265 223 402 206 ... 210 188 131 144
array([265, 223, 402, ..., 188, 131, 144])
-
-
-
win(replay)float641.0 1.0 0.0 0.0 ... 0.0 0.0 0.0 0.0
array([1., 1., 0., 0., 1., 0., 1., 0., 1., 0., 0., 0., 1., 1., 0., 0., 0., 1., 0., 1., 1., 1., 1., 1., 0., 0., 1., 0., 0., 0., 1., 1., 0., 1., 1., 1., 0., 0., 1., 1., 1., 1., 1., 1., 0., 1., 1., 0., 0., 0., 0., 0., 0., 1., 0., 1., 1., 1., 0., 0., 1., 1., 1., 1., 1., 0., 0., 0., 1., 0., 0., 1., 1., 0., 1., 1., 1., 0., 0., 0., 0., 0., 1., 0., 1., 0., 0., 0., 0., 1., 1., 1., 0., 0., 1., 1., 0., 1., 1., 0., 0., 0., 0., 1., 1., 1., 0., 0., 1., 1., 0., 1., 0., 0., 1., 0., 1., 1., 0., 0., 0., 1., 0., 0., 0., 1., 0., 1., 0., 0., 0., 0., 1., 1., 1., 1., 1., 0., 1., 0., 0., 0., 1., 1., 0., 1., 0., 1., 0., 1., 0., 1., 0., 0., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 0., 0., 0., 0., 0., 0., 0., 1., 0., 1., 1., 0., 0., 0., 1., 1., 1., 1., 1., 0., 1., 1., 0., 0., 1., 1., 1., 1., 1., 0., 1., 0., 0., 0., 0., 0., 0.])
-
-
- created_at :
- 2020-06-20T14:41:40.233660
- arviz_version :
- 0.8.3
- inference_library :
- pymc3
- inference_library_version :
- 3.9.1
xarray.Dataset
-
-
-
- replay: 202
-
-
replay(replay)int64265 223 402 206 ... 210 188 131 144
array([265, 223, 402, ..., 188, 131, 144])
-
-
-
enemy_race(replay)int320 1 0 2 0 1 0 1 ... 0 2 1 2 0 1 0 1
array([0, 1, 0, 2, 0, 1, 0, 1, 1, 2, 0, 1, 1, 2, 2, 1, 0, 1, 1, 0, 2, 0, 0, 0, 2, 2, 0, 2, 2, 0, 2, 2, 2, 1, 1, 0, 2, 2, 1, 1, 2, 1, 1, 1, 0, 0, 0, 2, 1, 2, 2, 0, 0, 0, 2, 2, 2, 2, 0, 0, 2, 0, 0, 0, 0, 0, 2, 0, 2, 2, 1, 2, 2, 0, 1, 1, 1, 2, 2, 0, 1, 1, 2, 1, 2, 0, 1, 1, 1, 1, 0, 0, 0, 2, 0, 0, 1, 1, 0, 0, 0, 1, 2, 0, 2, 1, 2, 2, 0, 0, 0, 0, 1, 2, 2, 0, 0, 0, 1, 1, 1, 2, 2, 2, 1, 0, 2, 0, 2, 0, 2, 2, 2, 1, 2, 2, 2, 0, 0, 2, 2, 1, 2, 1, 1, 1, 2, 0, 2, 1, 2, 2, 2, 1, 0, 0, 1, 2, 2, 1, 2, 0, 0, 1, 2, 0, 0, 0, 2, 2, 0, 2, 2, 2, 1, 0, 1, 1, 1, 0, 0, 0, 0, 2, 0, 2, 0, 2, 2, 1, 2, 0, 1, 2, 0, 2, 1, 2, 0, 1, 0, 1], dtype=int32)
-
enemy_mmr(replay)int324038 4163 4089 ... 4185 4000 4188
array([4038, 4163, 4089, 4196, 4005, 4043, 4036, 4078, 3928, 3936, 3988, 4014, 4028, 3915, 3896, 4017, 4000, 3936, 4037, 3845, 3980, 3889, 3979, 3997, 4075, 4115, 3928, 4092, 4006, 3995, 4034, 3889, 3900, 4011, 3962, 3955, 4054, 3959, 3879, 4072, 4032, 3986, 3966, 4002, 4238, 3995, 4225, 4098, 4451, 4135, 4042, 4008, 4011, 4152, 4108, 4095, 4071, 4337, 3848, 4111, 4009, 4144, 4021, 4136, 4213, 4401, 4444, 4467, 4060, 4193, 4175, 3999, 4004, 4285, 4019, 4190, 4166, 4183, 4076, 4242, 4017, 4228, 4001, 4099, 4050, 3973, 4037, 3905, 4240, 4065, 3975, 3919, 4091, 4093, 4045, 4045, 4101, 3975, 4189, 4047, 4003, 4040, 3997, 4010, 4023, 4004, 3939, 4153, 3962, 4030, 4050, 4080, 4093, 4071, 4013, 4146, 4109, 4019, 4049, 3998, 4021, 3959, 3985, 3990, 4002, 4033, 3883, 4123, 4199, 3851, 3889, 4045, 5035, 3856, 3924, 3902, 4088, 3921, 4055, 3952, 4027, 3905, 3770, 3940, 3899, 4082, 3921, 3931, 3959, 3998, 3888, 4039, 3988, 4025, 3831, 3952, 3945, 3869, 3981, 3833, 4008, 4119, 4072, 4047, 4018, 4427, 4105, 4127, 4225, 4142, 3944, 4040, 4016, 4126, 3960, 3981, 4040, 4074, 4093, 3887, 4173, 4103, 3929, 4040, 4081, 3949, 4123, 4253, 4223, 4059, 4028, 3946, 4244, 4304, 4182, 4370, 4194, 4147, 4232, 4185, 4000, 4188], dtype=int32)
-
-
- created_at :
- 2020-06-20T14:41:40.235138
- arviz_version :
- 0.8.3
- inference_library :
- pymc3
- inference_library_version :
- 3.9.1
xarray.Dataset
-
az.plot_trace(output, var_names=var_names);
az.plot_posterior(output, var_names = var_names);
az.summary(output, var_names = var_names)
mean | sd | hdi_3% | hdi_97% | mcse_mean | mcse_sd | ess_mean | ess_sd | ess_bulk | ess_tail | r_hat | |
---|---|---|---|---|---|---|---|---|---|---|---|
μ[0] | 4187.331 | 93.433 | 4014.681 | 4370.002 | 0.827 | 0.588 | 12761.0 | 12620.0 | 12833.0 | 5208.0 | 1.0 |
μ[1] | 4014.234 | 97.855 | 3835.751 | 4202.949 | 0.837 | 0.592 | 13669.0 | 13643.0 | 13703.0 | 5413.0 | 1.0 |
μ[2] | 3961.245 | 88.126 | 3794.804 | 4122.579 | 0.716 | 0.516 | 15147.0 | 14584.0 | 15149.0 | 5685.0 | 1.0 |
σ[0] | 79.342 | 60.635 | 0.027 | 187.229 | 0.751 | 0.531 | 6516.0 | 6516.0 | 5236.0 | 3992.0 | 1.0 |
σ[1] | 80.221 | 59.112 | 0.003 | 184.095 | 0.694 | 0.491 | 7250.0 | 7250.0 | 5667.0 | 4667.0 | 1.0 |
σ[2] | 82.427 | 62.380 | 0.020 | 196.670 | 0.784 | 0.554 | 6330.0 | 6330.0 | 5219.0 | 4182.0 | 1.0 |
And this seems to show I've grown more balanced as a player - the daily fluctuations are much more similar between games, I've gotten much better against Protoss, slightly worse against Terran, and my PvZ looks like it'll need some more work. Still, I'm really happy to see the data show I'm not as bad in PvP now!
For the next post, I'm torn between a few experiments I'm running:
- a hierarchical model to estimate both true global and per-matchup MMR at the same time while sharing information between the three matchups
- map dependence, which is going to be easy-ish now that I know how to index variables well
- actual time dependence, but I'll have to read up more on Gaussian processes and random walks to do that.
- playing around with prior- and posterior- predictive checks, which was going to be in this post, but it turns out I don't fully understand them enough yet. This will involve a foray into xarray - the awesome data structure that stores our results.
If you have any preference, please say so in the comments; I'll try to take it into account. Until the next time!
Comments