def test_threshold():
X, y = iris_data()
ax, threshold, count = ecdf(x=X[:, 0],
x_label='sepal length (cm)',
percentile=0.8)
assert threshold == 6.5
assert count == 120
def test_threshold():
X, y = iris_data()
ax, threshold, count = ecdf(x=X[:, 0],
x_label='sepal length (cm)',
percentile=0.8)
assert threshold == 6.5
assert count == 120
def plot_ecdf(self, result_max):
plt.clf()
axis = None
strategy_icon = ['1', '|', '_', '.', 'o', '2']
for indx, result in enumerate(result_max):
ax, _, _ = ecdf(x=result, ecdf_marker=strategy_icon[indx])
axis = ax
plt.legend([
'Fixed 1', 'Fixed 2', 'Fixed 4', 'Fixed 8', 'Fixed 12', 'MARTINI'
],
loc='upper center',
ncol=3,
fancybox=True,
bbox_to_anchor=(0.5, 1.15))
plt.xlabel('Time Interval (minutes)')
plt.savefig('metric_max_ecdf.pdf', bbox_inches="tight", format='pdf')
# check unique -not repeated- labels/item in the data-set
#summary_stats_unique = raw_data_frame.nunique().T
# check data types
#data_frame_data_types = raw_data_frame.info()
###############################################################################
## EXPLORATORY DATA ANALYSIS (EDA): Empirical Cumulative Density Function (ECDF)
# close picture in order to avoid overwriting with previous pics
plt.clf()
# set ECDF
ax, _, _ = ecdf(x=df_A.loc[:, 'RE'],
x_label='relative engagement',
ecdf_color='green')
ax, _, _ = ecdf(x=df_B.loc[:, 'RE'], ecdf_color='red')
#ax, _, _ = ecdf(x = test_c, ecdf_color = 'yellow')
#ax, _, _ = ecdf(x = merged_df_percentile_2.loc[:, 'percentiles'], ecdf_color = 'red')
#ax, _, _ = ecdf(x = merged_df_percentile_3.loc[:, 'percentiles'], ecdf_color = 'blue')
#ax, _, _ = ecdf(x = merged_df_percentile.loc[:, 'percentiles'], ecdf_color = 'yellow')
handles, labels = ax.get_legend_handles_labels()
ax.legend(
handles,
labels=['test_a', 'test_b'],
#labels = ['RE_100_quantiles', 'RE_1000_quantiles', 'RE_log', 'RE_log_10'],
framealpha=0.3,
scatterpoints=1,
loc='upper left')