In [1]:

    

from __future__ import division
import pymongo, pandas, random
import numpy as np
import matplotlib as mpl
import matplotlib.pyplot as plt
from scipy import stats

%matplotlib inline
plt.rcdefaults()
mpl.style.use('ggplot')

connection = pymongo.MongoClient('localhost', 27017)

# Top5 comunidades em número de usuários na época do dump
# communities = ["stackoverflow", "superuser", "serverfault", "math", "programmers"]
communities = ['ux']


    

In [2]:

    

def plot_histogram(female_sample, male_sample, place, name):
    bins=range(10)
    place.hist(np.log(np.array(female_sample) +1),bins, normed=True, label="females",alpha=0.5, color=mpl.cm.Dark2(0))
    place.hist(np.log(np.array(male_sample)+1),bins, normed=True, label="males", alpha=0.5)
    place.set_ylim(ymax=1.2)
    legend=place.legend(loc='best')
    frame = legend.get_frame()
    frame.set_facecolor('white')
    frame.set_edgecolor('grey')
    place.set_title(name, fontsize=13, fontweight='bold')
    place.grid(color='grey', linestyle='-.', linewidth=0.5, alpha=0.7, zorder=0, which="both")
    place.set_axis_bgcolor('white')


    

In [3]:

    

def plot_histogram2(female_sample, male_sample, place, name):
    data = np.hstack([list(female_sample.dropna()), list(male_sample.dropna())])
    bins = np.histogram(data)[1]
    place.hist(np.log(np.array(female_sample) +1),bins, normed=True, label="females",alpha=0.5, color=mpl.cm.Dark2(0))
    place.hist(np.log(np.array(male_sample)+1),bins, normed=True, label="males", alpha=0.5)
    legend=place.legend(loc='best')
    frame = legend.get_frame()
    frame.set_facecolor('white')
    frame.set_edgecolor('grey')
    place.set_title(name, fontsize=13, fontweight='bold')
    place.grid(color='grey', linestyle='-.', linewidth=0.5, alpha=0.7, zorder=0, which="both")
    place.set_axis_bgcolor('white')


    

In [4]:

    

def plot_density(female_sample, male_sample, place, name):
    males_ = male_sample.dropna()
    females_ = female_sample.dropna()
    
    density_female = stats.kde.gaussian_kde(females_)
    density_male = stats.kde.gaussian_kde(males_)
    minimum = min(min(females_),min(males_))
    maximum = max(max(females_),max(males_))
    
    x = np.arange(minimum-1, maximum, .1)
    place.plot(x, density_female(x), label="females", color=mpl.cm.Dark2(0))
    place.plot(x, density_male(x), label="males")
    
    legend=place.legend(loc='best')
    frame = legend.get_frame()
    frame.set_facecolor('white')
    frame.set_edgecolor('grey')
    place.set_title(name, fontsize=13, fontweight='bold')
    place.grid(color='grey', linestyle='-.', linewidth=0.5, alpha=0.7, zorder=0, which="both")
    place.set_axis_bgcolor('white')


    

In [5]:

    

def plot_cumulative(female_sample, male_sample, place, name):
    values, base = np.histogram(female_sample)
    cumulative = np.cumsum(values)

    values2, base2 = np.histogram(male_sample)
    cumulative2 = np.cumsum(values2)

    place.set_xlim(xmin=-1, xmax=max(max(base), max(base2)))
    place.step(base[:-1], cumulative, label="females", color=mpl.cm.Dark2(0))
    place.step(base2[:-1], cumulative2, label="males")
    
    legend=place.legend(loc='best')
    frame = legend.get_frame()
    frame.set_facecolor('white')
    frame.set_edgecolor('grey')
    place.set_title(name, fontsize=13, fontweight='bold')
    place.grid(color='grey', linestyle='-.', linewidth=0.5, alpha=0.7, zorder=0, which="both")
    place.set_axis_bgcolor('white')


    

In [6]:

    

def plot_regular_histogram(female_sample, male_sample, place, name):
#     bins=range(10)
    place.hist(np.array(female_sample), label="females",alpha=0.5, color=mpl.cm.Dark2(0))
    place.hist(np.array(male_sample), label="males", alpha=0.5)
#     place.set_ylim(ymax=1.2)
    legend=place.legend(loc='best')
    frame = legend.get_frame()
    frame.set_facecolor('white')
    frame.set_edgecolor('grey')
    place.set_title(name, fontsize=13, fontweight='bold')
    place.grid(color='grey', linestyle='-.', linewidth=0.5, alpha=0.7, zorder=0, which="both")
    place.set_axis_bgcolor('white')


    

In [7]:

    

for idx, community in enumerate(communities):
    
    community_db = connection[community]['statistics']
    cursor = community_db.find({'contributions_total': {'$gt':0}, 'gender': {'$ne': "Unknown"}},
                       {u'_id': False, u'gender':True,
                       'questions_total':True,'answers_total':True,'comments_total':True,'contributions_total':True,
                       'accepted_rate':True,'mean_utility':True,'questions_avg':True,
                       'lifetime':True,'activity_freq':True,})
    
    fig, axes = plt.subplots(nrows=4, ncols=2, figsize=(10, 15), dpi=300)
    fig.suptitle(community, fontsize=15, fontweight='bold')
    fig.tight_layout()
    fig.subplots_adjust(hspace=.4, wspace=0.3, top=0.93)

    df =  pandas.DataFrame(list(cursor))
    females = df.query("gender == 'Female'")
    males = df.query("gender == 'Male'")
    
    plot_histogram(females['questions_total'], males['questions_total'], axes[0][0], u"Número de Perguntas")
    plot_histogram(females['answers_total'], males['answers_total'], axes[0][1], u"Número de Respostas")
    plot_histogram(females['comments_total'], males['comments_total'], axes[1][0], u"Número de Comentários")
    plot_histogram(females['contributions_total'], males['contributions_total'], axes[1][1], u"Número de Contribuições")
    plot_density(females['accepted_rate'], males['accepted_rate'], axes[2][0], u"Taxa de Aceitação")
    plot_density(females['mean_utility'], males['mean_utility'], axes[2][1], u"Utilidade Média")
    plot_density(females['questions_avg'], males['questions_avg'], axes[3][0], u"Média dos Votos das Perguntas")
    plot_density(females['activity_freq'], males['activity_freq'], axes[3][1], u"Frequência de Atividade")
#     plot_histogram(females['lifetime'], males['lifetime'], axes[3][1], "Tempo de Vida")

    plt.savefig("test/"+community+".pdf", format="pdf")
    plt.close()


    

In [ ]: