forked from moreo/QuaPy
53 lines
1.4 KiB
Python
53 lines
1.4 KiB
Python
import gzip
|
|
import os
|
|
import sys
|
|
from collections import Counter
|
|
from Ordinal.utils import jaggedness
|
|
import pickle
|
|
import numpy as np
|
|
|
|
amazon = np.genfromtxt('prevalence_votes1_reviews100.csv', delimiter='\t')
|
|
telescope = np.genfromtxt('fact_real_prevalences.csv', delimiter=',')[1:]
|
|
|
|
nclasses_amazon = amazon.shape[1]
|
|
nclasses_telescope = telescope.shape[1]
|
|
|
|
jags_amazon = np.asarray([jaggedness(p) for p in amazon])
|
|
jags_telescope = np.asarray([jaggedness(p) for p in telescope])
|
|
|
|
import matplotlib.pyplot as plt
|
|
from matplotlib.pyplot import figure
|
|
import seaborn as sns
|
|
|
|
sns.set_theme('paper')
|
|
sns.set_style('dark')
|
|
sns.set(font_scale=0.7)
|
|
|
|
# figure, axis = plt.subplots(1, 2, figsize=(8, 7))
|
|
ymax = 0.75
|
|
|
|
figure(figsize=(8, 4), dpi=300)
|
|
|
|
ax=plt.subplot(1, 2, 1)
|
|
classes = np.arange(1, nclasses_amazon+1)
|
|
plt.bar(classes, np.mean(amazon, axis=0), yerr=np.std(amazon, axis=0), width=1)
|
|
ax.set_ylim(0, ymax)
|
|
ax.set_xlabel("stars")
|
|
ax.set_xticks(classes)
|
|
ax.set_title(f'Amazon Books ({jags_amazon.mean():.4f})')
|
|
|
|
ax=plt.subplot(1, 2, 2)
|
|
# ax=plt.subplot(1, 1, 1)
|
|
classes = np.arange(1, nclasses_telescope+1)
|
|
plt.bar(classes, np.mean(telescope, axis=0), yerr=np.std(telescope, axis=0), width=1)
|
|
ax.set_ylim(0, ymax)
|
|
ax.set_xlabel("energy bin")
|
|
ax.set_xticks(classes)
|
|
ax.set_title(f'FACT Samples ({jags_telescope.mean():.4f})')
|
|
|
|
|
|
plt.subplots_adjust(wspace=0.1, hspace=0)
|
|
plt.savefig('prevalence_averages.pdf', bbox_inches='tight')
|
|
|
|
|