def plot_histogram_with_dummies_theoretical(self):
if not self.clustering:
raise ValueError('Clustering not performed yet')
new_histogram, full_histogram = utils.build_clustered_histograms(
self.histogram, self.labels)
utils.plot_histograms(original_histogram=new_histogram,
theoretical_dummies=full_histogram)
def plot_histogram_with_dummies_real(self):
if not self.dummies:
raise ValueError('Dummies not generated yet')
_, theoretical_dummies = utils.build_clustered_histograms(
self.histogram, self.labels)
histogram_with_dummies = self.patient_concepts_matrix_dummies.sum(0)
cluster_indexing = theoretical_dummies.index
utils.plot_histograms(self.histogram[cluster_indexing],
theoretical_dummies[cluster_indexing],
histogram_with_dummies[cluster_indexing])
# print(data.head())
data = data[data['Year'] < 2002]
data['RD'] = data['RS'] - data['RA']
in_playoffs = data[data['Playoffs'] == 1]
out_playoffs = data[data['Playoffs'] == 0]
from utils import plot_scatters, plot_histograms, plot_deriving_slope
plot_scatters(in_playoffs, out_playoffs, label='RS')
plot_scatters(in_playoffs,
out_playoffs,
label='RD',
RDxline=False,
x=data['RD'].values,
y=data['W'].values)
plot_histograms(data)
# plot_deriving_slope(data['RD'].values, data['W'].values)
from matplotlib import pyplot as plt
plt.show()
rd_to_wins = data[['RD', 'W']].corr()
dePodesta = data[['OBP', 'SLG', 'BA', 'RS']].corr()
from sklearn.linear_model import LinearRegression
modelRS = LinearRegression()
modelRS.fit(data[['OBP', 'SLG']].values, data['RS'].values)
print(f"bias RS: {modelRS.intercept_}")
print(f"Coefficients: {modelRS.coef_}")
RA_data = data.dropna()
def plot_histogram(self):
utils.plot_histograms(original_histogram=self.histogram)
plt.figure(3)
plt.plot(np.arange(T), opt_arms_rate_mean, label=algo)
plt.fill_between(time,
opt_arms_rate_mean -
(q / np.sqrt(n_itr)) * np.sqrt(opt_arms_rate_var),
opt_arms_rate_mean +
(q / np.sqrt(n_itr)) * np.sqrt(opt_arms_rate_var),
color='#D3D3D3')
plt.xlabel('Rounds')
plt.ylabel('$\%$')
plt.title('Rate of optimal arms pulling, averaged over ' + str(n_itr) +
' runs')
plt.legend()
if plot_histo:
fig = plot_histograms(algo, histograms, hist_times, K, A_star)
alpha_star, opt_mix_rew = MO_MAB.alpha_star, MO_MAB.optimal_mixed_rew
print('Optimal mixed reward = ' + str(opt_mix_rew))
alpha = ogde.alpha.reshape((1, K))
opt_mix = alpha.dot(MO_MAB.O)[0]
print('Mixed reward at time T = ' + str(opt_mix))
print('')
print('Alpha_star = ' + str(alpha_star))
print('Alpha_T = ' + str(alpha[0]))
if plot_arms and D in [2, 3]:
plot_momab(MO_MAB,
opt_mix,
alpha_ogde=alpha_ogde,
print("You didn't specify any plot types, exiting")
sys.exit(1)
connectstring = dbutils.make_connectstring(db="views",
hostname="VIEWSHOST",
port="5432",
prefix="postgres",
uname="VIEWSADMIN")
dir_descriptive = "/storage/runs/current/descriptive"
dir_table = "/".join([dir_descriptive, schema, table])
df = dbutils.db_to_df(connectstring, schema, table, ids=[timevar, groupvar])
df.sort_index(inplace=True)
if plot_wawa:
utils.plot_world_average_with_actuals(df, connectstring, dir_table,
timevar, groupvar)
if plot_spaghetti:
utils.plot_spaghetties(df, connectstring, dir_table)
if plot_hist:
utils.plot_histograms(df, dir_table)
if plot_lpg:
utils.plot_lines_per_group(df, dir_table)
if plot_abt:
utils.plot_stats_by_time(df, dir_table)
if plot_pgcm:
utils.plot_pgcm(df, connectstring, dir_table)
if plot_lpgwa:
utils.plot_lines_per_group_with_actuals(df, connectstring, dir_table)