Number of page views each day at freeCodeCamp.org forum

Created by Darío López Díaz. Work in progress..

In this project we will analize the page views from FCC forums, we will visualize the data by year and month using Matplotlib and Seaborn packages. The idea is to detect the months with maximal number of views.

import matplotlib.pyplot as plt import pandas as pd import numpy as np import seaborn as sns from pandas.plotting import register_matplotlib_converters register_matplotlib_converters()

df = pd.read_csv('fcc-forum-pageviews.csv') df_dateindex = df.set_index('date')

We perform a 'filtering out' of the days with page views above the top 2.5% of the dataset or bottom 2.5% of the dataset. This filtering out is done in order to avoid statistical errors in data.

df_filtered = df_dateindex[(df_dateindex['value'] > df_dateindex['value'].quantile(0.025)) & (df_dateindex['value'] < df_dateindex['value'].quantile(0.975))]

The 'lineplot' of the 'page views vs date' of the filtered data is:

plt.figure(figsize=(19, 6)) line_plot = plt.plot(df_filtered) plt.xticks(np.arange(np.argmin(df_filtered), np.argmax(df_filtered), step=152)) plt.title('Daily freeCodeCamp Forum Page Views 5/2016-12/2019') plt.xlabel('Date') plt.ylabel('Page Views') plt.show()

We modify the DataFrame including two new columns by 'Year' and 'Month':  

df_filtered['year'] = pd.DatetimeIndex(df_filtered.index).year df_filtered['month'] = pd.DatetimeIndex(df_filtered.index).month_name()

Now we can plot the 'Page Views' by Months and Years

plt.figure(figsize=(8.7, 6)) sns.barplot(x='year', y='value', hue='month', data=df_filtered , hue_order = ['January', 'February', 'March', 'April', 'May', 'June', 'July', 'August', 'September', 'October', 'November', 'December'])

Now, to see wich is the month with more views, we can create a box plot, with the accumulative values along all the years.

f, axes = plt.subplots(1, 2, figsize=(16,7)) by_year = sns.boxplot(y='value', x= 'year', data=df_filtered, orient='v' , ax=axes[0], palette="Blues_d") by_month = sns.boxplot(y='value', x= 'month', data=df_filtered, orient='v' , ax=axes[1], palette="Blues_d") by_month.set_xticklabels(by_month.get_xticklabels(), rotation=45, horizontalalignment='right')

October and February are the months with more accumulative views along all the data.

.css-9i38tn{letter-spacing:inherit;-webkit-text-decoration:none;text-decoration:none;}Number of page views each day at freeCodeCamp.org forum

Number of page views each day at freeCodeCamp.org forum