import numpy as np
from statsmodels import api as sm
from sklearn.model_selection import StratifiedKFold
from sklearn.metrics import log_loss
from sklearn.preprocessing import LabelBinarizer
from statcast.bip import Bip
from statcast.better.sm import BetterGLM, BetterMNLogit
from statcast.better.spark import cross_val_predict
from statcast.tools.plot import plotPrecRec, plotPrecRecMN, plotResiduals
# %%
bip = Bip(years=(2016, ), n_jobs=-1)
# %%
xLabels = ['hit_speed', 'hit_angle', 'sprayAngle']
fancyLabels = ['Exit Velocity', 'Launch Angle', 'Spray Angle']
units = ['mph', 'degrees', 'degrees']
yLabel = 'events'
subData = bip.data.loc[~bip.data['exclude'], xLabels + [yLabel]]
outs = [
'Bunt Groundout', 'Double Play', 'Fielders Choice', 'Fielders Choice Out',
'Flyout', 'Forceout', 'Grounded Into DP', 'Groundout', 'Lineout',
'Pop Out', 'Runner Out', 'Sac Bunt', 'Sac Fly', 'Sac Fly DP',
'Triple Play', 'Bunt Pop Out', 'Bunt Lineout', 'Sacrifice Bunt DP'
from matplotlib import pyplot as plt
from statcast.bip import Bip
from statcast.tools.plot import correlationPlot
from statcast.better.utils import findTrainSplit
# %% Plot correlation of imputing model
years = (2016, 2015)
labels = ['Exit Velocity', 'Launch Angle', 'Hit Distance']
units = ['mph', 'degrees', 'feet']
for year in years:
bip = Bip(years=(year, ), n_jobs=-1)
testData = bip.data.loc[~bip.data.exclude & ~bip.data.scImputed, :]
testY = bip.scImputer.createY(testData)
testYp = bip.scImputer.predictD(testData)
labelsYr = ['{} {}'.format(label, year) for label in labels]
figs = correlationPlot(testY, testYp, labels=labelsYr, units=units, ms=0.7)
for fig, label in zip(figs, labels):
fig.savefig('{} Correlation {}'.format(label, year))
# %% Plot Tree Curve
# %% Imports
from scipy import stats
from matplotlib import pyplot as plt
from statcast.bip import Bip
from statcast.plot import plotMLBLogos
from statcast.tools.plot import addText
# %%
bip15 = Bip(years=(2015, ), n_jobs=-1)
bip16 = Bip(years=(2016, ), n_jobs=-1)
# %% Plot Correlations
labels = ['hit_speed', 'hit_angle', 'hit_distance_sc']
units = ['mph', 'degrees', 'feet']
fancyLabels = ['Exit Velocity', 'Launch Angle', 'Hit Distance']
for i, (label, unit, fancyLabel) in enumerate(zip(labels, units, fancyLabels)):
if '(scImputed||home_team)' in bip15.scFactorMdl.formulas[i]:
x = bip15.scFactorMdl.factors_[label]['home_team']['(Intercept)'] + \
bip15.scFactorMdl.factors_[label]['home_team']['scImputedFALSE']
missing15 = False
else:
x = bip15.scFactorMdl.factors_[label]['home_team']['(Intercept)']
missing15 = True
if '(scImputed||home_team)' in bip16.scFactorMdl.formulas[i]:
y = bip16.scFactorMdl.factors_[label]['home_team']['(Intercept)'] + \
bip16.scFactorMdl.factors_[label]['home_team']['scImputedFALSE']
import datetime
import requests
from pyspark import SparkContext
from statcast.bip import Bip
# %% Create Spark Context
sc = SparkContext(appName="post5")
# %% Load data, plot histograms of statcast data
years = (2015, 2016)
for year in years:
bip = Bip(years=(year, ), n_jobs=sc)
bip.plotSCHistograms()
# %% Transfer results to S3
instanceID = requests. \
get('http://169.254.169.254/latest/meta-data/instance-id').text
dtStr = datetime.datetime.utcnow().strftime('%Y-%m-%d--%H-%M-%S')
os.system('aws s3 sync . s3://mf-first-bucket/output/{}/{}'.format(
instanceID, dtStr))
# %% Stop Spark Context
sc.stop()