
            
              import re
import csv
from collections import Counter

import numpy as np
import requests
import matplotlib.pyplot as plt
import seaborn

raw_data_str = requests.get('https://raw.githubusercontent.com/TheSCInitiative/BOINC/main/BOINC_Census/2022/data_source/BOINC_Census_2022_results_raw_cleaned.csv').text
data_str = requests.get('https://raw.githubusercontent.com/TheSCInitiative/BOINC/main/BOINC_Census/2022/data_source/BOINC_Census_2022_results_completed.csv').text
raw_data = [row for row in csv.reader(raw_data_str.split('\n')) if bool(row)]
data = [row for row in csv.reader(data_str.split('\n')) if bool(row)]
headers = data[0]

# Remove headers from data
raw_data = raw_data[1:]
data = data[1:]

# Eye candy settings
BINARY_COLOURS = ['#38CF62', '#E65050']
GENERAL_COLOURS = seaborn.color_palette("colorblind")
BOINC_BLUE = '#163E72'

def get_col(heading: str, data_source: list):
    idx = headers.index(heading)
    return [row[idx] for row in data_source]

def multi_sel_to_hist(mult_sel_col: list[str]):
    '''
    Converts a column with CSV values in each cell to a simple list that can
    be turned into a histogram.
    
    Fillout does not perform escaping on internal commas so we have to handle it.
    Because I always put a space after a comma, we can safely replace it.
    '''
    return sum([row.replace(', ', ' ').split(',') for row in mult_sel_col], start=[])


            
              print(f'Total Submissions: {len(raw_data)}')
print(f'Completed Submissions: {len(data)}')

fig, ax = plt.subplots(1, 2)
fig.set_figwidth(15)

used_boinc_counter = Counter(get_col('Do you or have you ever used BOINC?', data))

ax[0].pie([len(data), len(raw_data) - len(data)], labels=['Completed', 'Not Completed'], autopct='%.0f%%', colors=BINARY_COLOURS)
ax[1].pie([used_boinc_counter['Yes'], used_boinc_counter['No']], labels=['Yes', 'No'], autopct='%.0f%%', colors=BINARY_COLOURS)
ax[0].title.set_text('Form Completion Rate')
ax[1].title.set_text('Do you or have you ever used BOINC?')
plt.show()

Total Submissions: 1119
Completed Submissions: 939


            
              fig, ax = plt.subplots(1, 2)
fig.set_figwidth(15)
gender_groups = Counter(get_col('What gender do you identify as?', data)).items()
age_groups = Counter(get_col('What is your age?', data))
age_groups_sorted = sorted(age_groups.items(), key=lambda v: ord(v[0][0]) + 999 * int(v[0].endswith('+'))) # Ensure that 'more than' values are put to the end

ax[0].bar([a[0] for a in gender_groups], [a[1] for a in gender_groups], color=BOINC_BLUE)
ax[1].bar([a[0] for a in age_groups_sorted], [a[1] for a in age_groups_sorted], color=BOINC_BLUE)
ax[0].title.set_text('What gender do you identify as?')
ax[1].title.set_text('What is your age?')

plt.show()


            
              fig, ax = plt.subplots(1, 2)
fig.set_figwidth(15)
edu_groups = Counter(get_col('What is your highest level of education you have completed?', data))
country_groups = Counter(get_col('What country are you from?', data))
# Take the top 5 sources to prevent clutter
country_groups = sorted(country_groups.items(), key=lambda x: x[1], reverse=True)[:5]
# Combine none and other
edu_groups['None or Other'] = edu_groups['No Education'] + edu_groups['Other']
edu_groups.pop('No Education')
edu_groups.pop('Other')

repl_dict = {
    'United States of America' : 'US',
    'United Kingdom' : 'UK',
    'Tertiary Education (postgraduate)' : 'Uni Postgrad', 
    'Tertiary Education (undergraduate)' : 'Uni Undergrad', 
    'Secondary Education' : 'Secondary', 
    'Primary Education' : 'Primary'
}

edu_groups = edu_groups.items()

ax[0].bar([repl_dict.get(a[0], a[0]) for a in edu_groups], [a[1] for a in edu_groups], color=BOINC_BLUE)
ax[1].bar([repl_dict.get(a[0], a[0]) for a in country_groups], [a[1] for a in country_groups], color=BOINC_BLUE)
ax[0].title.set_text('What is your highest level of education?')
ax[1].title.set_text('What country are you from? (top-five)')

plt.show()


            
              fig, ax = plt.subplots(1, 3)
fig.set_figwidth(15)
computers = Counter(get_col('How many computers do you have that run BOINC regularly?', data))
computers_sorted = sorted(computers.items(), key=lambda v: ord(v[0][0]) + 999 * int(v[0].endswith('+'))) # Ensure that 'more than' values are put to the end
mobiles = Counter(get_col('How many of the above computers are mobile devices? (mobile phones or tablets)', data))
mobiles_sorted = sorted(mobiles.items(), key=lambda v: ord(v[0][0]) + 999 * int(v[0].endswith('+'))) # Ensure that 'more than' values are put to the end

ax[0].bar([c[0] for c in computers_sorted], [c[1] for c in computers_sorted], color=BOINC_BLUE)
ax[1].bar([c[0] for c in mobiles_sorted], [c[1] for c in mobiles_sorted], color=BOINC_BLUE)
ax[2].pie([len(data) - mobiles['0'], mobiles['0']], labels=['BOINC on mobile', 'No BOINC mobile'], autopct='%.0f%%', colors=BINARY_COLOURS)
ax[0].title.set_text('How many devices do you have running BOINC?')
ax[1].title.set_text('How many are mobile devices?')
ax[2].title.set_text('BOINC on mobile rate')

plt.show()


            
              boinc_often = Counter(get_col('How often do you run BOINC?', data)).items()

repl_dict = {
    'A few days of the week' : 'Few days/wk',
    'All the time (24/7)' : '24/7',
    'Most days of the week' : 'Most days', 
    'A few hours every week' : '<1day/wk', 
    "I don't use BOINC" : 'N/A', 
    "Only when I'm not using the computer" : 'Only Idle'
}

plt.bar([repl_dict.get(a[0], a[0]) for a in boinc_often], [a[1] for a in boinc_often], color=BOINC_BLUE)
plt.title('How often do you run BOINC?')
plt.show()


            
              fig, ax = plt.subplots(1, 2)
fig.set_figwidth(15)
contrib_impact = Counter(get_col('How much impact do you think your BOINC contribution is having?', data))
contrib_impact.pop('') # Empty responses were recorded due to an error in the form. This occurred 13 times.
contrib_impact = contrib_impact.items()
energy_src = Counter(get_col('What source do you get most of your electricity from?', data)).items()

ax[0].bar([a[0] for a in contrib_impact], [a[1] for a in contrib_impact], color=BOINC_BLUE)
ax[1].bar([re.sub(r' \(.*\)', '', a[0]) for a in energy_src], [a[1] for a in energy_src], color=BOINC_BLUE)
ax[0].title.set_text('What impact do you think your contribution is having?')
ax[1].title.set_text('What source do you get most of your electricity from?')

plt.show()


            
              fig, ax = plt.subplots(1, 2)
fig.set_figwidth(15)

financially = list(Counter(get_col('Have you ever supported a BOINC project financially?', data)).items())
financially.sort(reverse=True)
non_financially = list(Counter(get_col('Aside from running BOINC, have you ever supported a BOINC project non-financially? (eg, donating computer hardware, volunteering time)', data)).items())
non_financially.sort(reverse=True)

ax[0].pie([a[1] for a in financially], labels=[a[0] for a in financially], autopct='%.0f%%', colors=BINARY_COLOURS)
ax[1].pie([a[1] for a in non_financially], labels=[a[0] for a in non_financially], autopct='%.0f%%', colors=BINARY_COLOURS)
ax[0].title.set_text('Supported a BOINC project financially')
ax[1].title.set_text('Supported a BOINC project non-financially')

plt.show()


            
              fig, ax = plt.subplots(1, 2)
fig.set_figwidth(15)
when_boinc = sorted(Counter(get_col('What year did you first hear about BOINC?', data)).items())
where_boinc = Counter(get_col('How did you first hear about BOINC?', data))
# Take the top 5 sources to prevent clutter
where_boinc_top = sorted(where_boinc.items(), key=lambda x: x[1], reverse=True)[:5]
repl_dict = {
    'Via a Google search or other search engine' : 'Web search',
    'From a news article (both print or digital)' : 'News',
    'Word-of-mouth/a friend told me' : 'Friend'
}

ax[0].bar([a[0][2:] for a in when_boinc], [a[1] for a in when_boinc], color=BOINC_BLUE)
ax[1].bar([repl_dict.get(a[0], a[0]) for a in where_boinc_top], [a[1] for a in where_boinc_top], color=BOINC_BLUE)
ax[0].title.set_text('What year did you first hear about BOINC?')
ax[1].title.set_text('How did you first hear about BOINC? (top-five)')

plt.show()


            
              fig, ax = plt.subplots(1, 2)
fig.set_figwidth(15)
where_help = Counter(multi_sel_to_hist(get_col('Where do you seek help for problems you have with running BOINC?', data)))
where_hear = Counter(multi_sel_to_hist(get_col('In what ways would you like to hear about what BOINC projects are doing?', data)))
# Take the top 5 options to prevent clutter
where_help_top = sorted(where_help.items(), key=lambda x: x[1], reverse=True)[:5]
where_hear_top = sorted(where_hear.items(), key=lambda x: x[1], reverse=True)[:5]

repl_dict = {
    'BOINC project forums/message boards' : 'Proj. forums',
    'Google search or other search engine' : 'Web search',
    'The official BOINC forum/message board' : 'BOINC forum',
    'The BOINC website' : 'BOINC website',
    'Chat groups like Discord or Telegram' : 'Chat groups',
    'Via the BOINC manager' : 'BOINC manager',
    'Via a web page or dashboard' : 'Web page',
    'Email notifications' : 'Email',
    'Forum/message board posts' : 'Forum post'
}

ax[0].bar([repl_dict.get(a[0], a[0]) for a in where_help_top], [a[1] for a in where_help_top], color=BOINC_BLUE)
ax[1].bar([repl_dict.get(a[0], a[0]) for a in where_hear_top], [a[1] for a in where_hear_top], color=BOINC_BLUE)
ax[0].title.set_text('Where do you seek help for BOINC? (top-five)')
ax[1].title.set_text('How do you want to hear about BOINC? (top-five)')

plt.show()

Results of the BOINC Census 2022¶

Introduction¶

Basic Statistics¶

Demographics¶

BOINC usage¶

BOINC Impact¶

BOINC Origins¶

Project Administration¶