import pandas as pd import numpy as np import seaborn as sns import matplotlib.pyplot as plt import warnings warnings.filterwarnings( 'ignore' ) sns.set(style='darkgrid') plt.rcParams["figure.figsize"] = (15,6)

def get_label(g): for p in g.patches: height = p.get_height() g.text(p.get_x()+p.get_width()/2., height/2, '{}'.format(round(height)), ha="center", color='white')

df_2019 = pd.read_csv("/work/multiple_choice_responses.csv") df_2020 = pd.read_csv("/work/kaggle_survey_2020_responses.csv") df_2021 = pd.read_csv("/work/kaggle_survey_2021_responses.csv")

# Select just the column that i will use df_2019 = df_2019[['Time from Start to Finish (seconds)', 'Q1', 'Q2', 'Q3', 'Q4', 'Q5']] df_2020 = df_2020[['Time from Start to Finish (seconds)', 'Q1', 'Q2', 'Q3', 'Q4', 'Q5']] df_2021 = df_2021[['Time from Start to Finish (seconds)', 'Q1', 'Q2', 'Q3', 'Q4', 'Q5']] # rename columns cols = {'Time from Start to Finish (seconds)':'Time_start_to_finish', 'Q1':'age', 'Q2':'gender', 'Q3':'country_reside', 'Q4':'formal_education', 'Q5':'title_job', 'Q6':'years_writing_code', } df_2019.rename(columns=cols, inplace=True) cols = {'Time from Start to Finish (seconds)':'Time_start_to_finish', 'Q1':'age', 'Q2':'gender', 'Q3':'country_reside', 'Q4':'formal_education', 'Q5':'title_job', 'Q6':'years_writing_code', } df_2020.rename(columns=cols, inplace=True) cols = {'Time from Start to Finish (seconds)':'Time_start_to_finish', 'Q1':'age', 'Q2':'gender', 'Q3':'country_reside', 'Q4':'formal_education', 'Q5':'title_job', 'Q6':'years_writing_code', } df_2021.rename(columns=cols, inplace=True) df_2019.drop(0, axis=0, inplace=True) df_2020.drop(0, axis=0, inplace=True) df_2021.drop(0, axis=0, inplace=True) # creare a new column to indentify what year the dataset belongs df_2019['year'] = '2019' df_2020['year'] = '2020' df_2021['year'] = '2021' aux = pd.concat([df_2019, df_2020], ignore_index=True) df_all = pd.concat([aux, df_2021], ignore_index=True) #Change the extense name df_all['country_reside'] = df_all['country_reside'].str.replace('Iran, Islamic Republic of...', 'Iran') df_all['country_reside'] = df_all['country_reside'].str.replace('United Kingdom of Great Britain and Northern Ireland', 'United Kingdom') df_all['country_reside'] = df_all['country_reside'].replace('Hong Kong (S.A.R.)', 'Hong Kong') # replace the gender df_all['gender'] = df_all['gender'].str.replace('Female', 'Woman') df_all['gender'] = df_all['gender'].str.replace('Male','Man') # Creare a new column continent map_continent = {'India':'Asia', 'Indonesia':'Asia', 'Pakistan':'Asia', 'Mexico':'North America', 'Russia':'Asia', 'Turkey':'Asia', 'Australia':'Australia', 'Nigeria':'Africa', 'Greece':'Europe', 'Belgium':'Europe', 'Japan':'Asia', 'Egypt':'Africa', 'Singapore':'Asia', 'Brazil':'South America', 'Poland':'Europe', 'China':'Asia','Iran':'Asia', 'United States of America':'North America', 'Italy':'Europe', 'Viet Nam':'Asia', 'Israel':'Asia', 'Peru':'South America', 'South Africa':'Africa', 'Other':'Other','Spain':'Europe', 'Bangladesh':'Asia', 'United Kingdom':'Europe', 'France':'Europe','Switzerland':'Europe', 'Algeria':'Africa', 'Tunisia':'Africa', 'Argentina':'South America', 'Sweden':'Europe', 'Colombia':'South America','I do not wish to disclose my location':'I do not wish to disclose my location', 'Canada':'North America','Chile':'South America', 'Netherlands':'Europe', 'Ukraine':'Europe', 'Saudi Arabia':'Asia', 'Romania':'Europe','Morocco':'Africa', 'Austria':'Europe', 'Taiwan':'Asia', 'Kenya':'Africa', 'Belarus':'Europe', 'Ireland':'Europe','Portugal':'Europe', 'Hong Kong':'Asia', 'Denmark':'Europe', 'Germany':'Europe','South Korea':'Asia', 'Philippines':'Asia', 'Sri Lanka':'Asia', 'United Arab Emirates':'Asia','Uganda':'Africa', 'Ghana':'Africa', 'Malaysia':'Asia', 'Thailand':'Asia', 'Nepal':'Asia', 'Kazakhstan':'Asia','Ethiopia':'Africa', 'Iraq':'Asia', 'Ecuador':'South America', 'Norway':'Europe', 'Czech Republic':'Europe', 'Kenya':'Africa', 'Republic of Korea':'Asia', 'New Zealand':'Oceanian', 'Hungary':'Europe'} df_all['continent'] = df_all['country_reside'].map(map_continent) # change types and trasnforming time ( seconds ) to minutes df_all['year'] = df_all['year'].astype(int) df_all['Time_start_to_finish'] = df_all['Time_start_to_finish'].astype(int) df_all['Time_start_to_finish'] = df_all['Time_start_to_finish'].apply(lambda x: round(x / 60, ndigits=2))

df_all.sample()

print(f'minimum time to complete the survey: {df_all["Time_start_to_finish"].min()} minutes') print(f'maximum time to complete the survey: {df_all["Time_start_to_finish"].max()} minutes')

df_year = df_all.groupby(['year'], as_index=False)['age'].count().rename( columns = {"age":'total'}) df_year = pd.DataFrame(df_year)

plt.figure(figsize=(15, 7)) g = sns.barplot(x='year', y='total', data = df_year, palette='Set1') plt.title('NUMBER OF SURVEY PARTICIPANTS PER YEAR') get_label(g)

df_age = df_all.groupby(['age', 'year'], as_index=False)['gender'].count().rename( columns = {"gender":'total'}) df_age = pd.DataFrame(df_age)

plt.figure(figsize=(15, 7)) g = sns.barplot(x='age', y='total', hue='year', data = df_age, palette='Set2') plt.title('AGE OF PARTICIPANTS PER YEAR');

mask = (df_all['gender'] != 'Prefer not to say') df_gender = df_all.loc[mask].groupby(['gender', 'year'], as_index=False)['age'].count().rename( columns = {'age':'total'})

plt.figure(figsize=(15, 7)) order = ['Man', 'Woman', 'Nonbinary', 'Prefer to self-describe'] g = sns.barplot(x='gender', y='total', hue='year', data = df_gender, order=order, palette='Set2') plt.title('GENDER OF PARTICIPANTS PER YEAR');

title_job = df_all.groupby(['title_job', 'year'], as_index=False)['age'].count().rename( columns = {'age':'total'})

plt.figure(figsize=(15, 7)) g = sns.barplot(x='title_job', y='total', hue='year', data = title_job, palette='Set2') plt.title('TITLE JOB OF PARTICIPANTS PER YEAR'); plt.xticks(rotation=90);

mask_country = (df_all['country_reside'] != 'Other') df_country = df_all.loc[mask_country].groupby(['country_reside', 'year'], as_index=False)['gender'].count().rename( columns = {"gender":'total'}) df_country = pd.DataFrame(df_country) df_country_10 = df_country.nlargest(30, 'total')

plt.figure(figsize=(15, 7)) g = sns.barplot(x='country_reside', y='total', hue='year', data = df_country_10, palette='magma') plt.title('COUNTRY OF PARTICIPANTS PER YEAR'); plt.xticks(rotation=90);

mask = (df_all['continent'] != 'Other') & (df_all['continent'] != 'I do not wish to disclose my location') df1_contnent = df_all.loc[mask].groupby(['continent', 'year'], as_index=False)['age'].count().rename( columns = {'age':'total'})

order = ['Asia', 'North America', 'Europe', 'Africa', 'South America', 'Oceanian'] g = sns.barplot(y='continent', x='total', hue='year', data = df1_contnent, order=order, palette='magma') plt.title('CONTINENT OF PARTICIPANTS PER YEAR');