NormalizedNumberOfCitationsPerStance.py

from matplotlib import pyplot as plt
from matplotlib import ticker as mtick
import seaborn as sns
import numpy as np
import pandas as pd
import json
from scipy.ndimage import gaussian_filter1d

sns.set_theme(style='whitegrid')

df = pd.concat([pd.read_json(r'../../Data/Model Predicted Data/%s-Predictions.json' % ds) for ds in ['NLP', 'ML']])
df = df[['url', 'year', 'stance', 'citations']].drop_duplicates()
data = df.to_dict(orient='records')

cp = sns.color_palette()[:8] + sns.color_palette('muted')[:8]

# create dataframe for sections of stance
min_x, max_x, N_x = -1, 1, 2*10+1
w_x = (max_x - min_x) / (N_x - 1)
df = pd.DataFrame([
	{
	'stance': next(x for x in np.linspace(min_x, max_x, N_x) if p['stance'] >= x-w_x/2 and p['stance'] < x+w_x/2),
	'citations': p['citations'],
	'year': p['year'],
	}
	for p in data
])

# calculate normalized citations
stats = df.groupby(['year'])['citations'].agg(['mean', 'std'])
df['normalized_citations'] = (df['citations'] - list(stats['mean'][df['year']])) / list(stats['std'][df['year']])

# create plot
stats = df.groupby(['stance'])['normalized_citations'].agg(['mean', 'count', 'std'])
ci95hi, ci95lo = list(), list()
for i in stats.index:
	m, c, s = stats.loc[i]
	ci95hi.append(m + 1.95*s/np.sqrt(c))
	ci95lo.append(m - 1.95*s/np.sqrt(c))
stats['ci95hi'] = ci95hi
stats['ci95lo'] = ci95lo
for c in stats.columns: stats[c] = gaussian_filter1d(stats[c], sigma=1.5)

plt.figure(figsize=(5.4, 4.0))
plt.hlines(0, -1, 1, linestyles='dotted', color='black')
plt.plot(stats['mean'], color='black')
plt.fill_between(stats.index, stats['ci95lo'], stats['ci95hi'], edgecolor='black', facecolor='black', alpha=0.06)

plt.title('Normalized Number of Citations per Stance')
plt.xlabel('Stance')
plt.ylabel('Normalized Number of Citations')
plt.subplots_adjust(bottom=0.14, left=0.15, right=0.98)
plt.savefig('NormalizedNumberOfCitationsPerStance.pdf')
plt.show()