Community¶

A summary of the demographic information of the NumPy survey respondents.

fname = "data/2021/numpy_survey_results.tsv"
column_names = [
    'age', 'gender', 'lang', 'lang_other', 'country', 'degree', 'degree_other',
    'field_of_study', 'field_other', 'role', 'role_other', 'version', 'version_other',
    'primary_use', 'share_code', 'programming_exp', 'numpy_exp', 'use_freq', 'components',
    'use_c_ext', 'prog_lang', 'prog_lang_other', 'surv2020'
]
demographics_dtype = np.dtype({
    "names": column_names,
    "formats": ['<U600'] * len(column_names),
})

data = np.loadtxt(
    fname, delimiter='\t', skiprows=3, dtype=demographics_dtype, 
    usecols=list(range(11, 33)) + [90], comments=None, encoding='UTF-16'
)

glue('2021_num_respondents', data.shape[0], display=False)

Demographics¶

Age¶

Of the 522 survey respondents, 456 (87%) shared their age.

The majority of respondents are in the age groups 25-34 and 35-44. Very few respondents are older than 55, and even fewer are younger than 18.

# Ignore empty fields and "prefer not to answer"
drop = np.logical_and(data['age'] != '', data['age'] != 'Prefer not to answer')
age = data['age'][drop]
labels, cnts = np.unique(age, return_counts=True)
ind = np.array([5,0,1,2,3,4])
labels, cnts = labels[ind], cnts[ind]

fig, ax = plt.subplots(figsize=(12, 8))
ax.bar(
  np.arange(len(labels)),
  100 * cnts / age.shape[0], 
  tick_label=labels,
)
ax.set_ylabel('Percentage of Respondents')
ax.set_xlabel('Age Group')
ax.set_title("Age Distribution of Survey Respondents");
fig.tight_layout()

glue('2021_num_age_respondents', gluval(age.shape[0], data.shape[0]), display=False)

Gender¶

Of the 522 survey respondents, 453 (87%) shared their gender.

An overwhelming majority of respondents identify as male. Only about 11% of respondents identify as female.

# Ignore empty fields and "prefer not to answer"
drop = np.logical_and(data['gender'] != '', data['gender'] != 'Prefer not to answer')
gender = data['gender'][drop]
labels, cnts = np.unique(gender, return_counts=True)

fig, ax = plt.subplots(figsize=(8, 8))
ax.pie(cnts, labels=labels, autopct='%1.1f%%')
ax.set_title("Gender Distribution");
fig.tight_layout()

glue('2021_num_gender', gluval(gender.shape[0], data.shape[0]), display=False)

Language Preference¶

Of the 522 respondents, 497 (95%) shared their preferred language.

Over 67% of respondents reported English as their preferred language.

# Ignore empty fields
lang = data['lang'][data['lang'] != '']
# Self-reported language
lang_other = data['lang_other'][data['lang_other'] != '']
lang_other = capitalize(lang_other)
lang = np.concatenate((lang, lang_other))
labels, cnts = np.unique(lang, return_counts=True)
cnts = 100 * cnts / cnts.sum()
I = np.argsort(cnts)[::-1]
labels, cnts = labels[I], cnts[I]

# Create a summary table
with open('_generated/language_preference_table.md', 'w') as of:
    of.write('| **Language** | **Preferred by % of Respondents** |\n')
    of.write('|--------------|-----------------------------------|\n')
    for lbl, percent in zip(labels, cnts):
        of.write(f'| {lbl} | {percent:1.1f} |\n')

glue('2021_num_lang_pref', gluval(lang.shape[0], data.shape[0]), display=False)

Click to show/hide table

Language	Preferred by % of Respondents
English	67.2
Other	6.6
Spanish	6.2
French	5.4
Japanese	5.0
Russian	1.8
German	1.4
Mandarin	1.0
Portuguese	0.6
Polish	0.6
Swedish	0.4
Czech	0.4
Italian	0.4
Turkish	0.4
Chinise	0.2
中文	0.2
Chinese	0.2
Bulgarian	0.2
Allemand	0.2
Hungarian	0.2
Dutch	0.2
Hebrew	0.2
Hindi	0.2
Indonesian	0.2
Korean	0.2
“norwegian “	0.2

Country of Residence¶

Of the 522 respondents, 440 (84%)shared their current country of residence. The survey saw respondents from 54 countries in all. A quarter of respondents reside in the United States.

The following chart shows the relative number of respondents from ~10 countries with the largest number of participants. For privacy reasons, countries with fewer than a certain number of respondents are not included in the figure, and are instead listed in the subsequent table.

# Preprocess data
country = data['country'][data['country'] != '']
country = strip(country)
# Distribution
labels, cnts = np.unique(country, return_counts=True)
# Privacy filter
num_resp = 10
cutoff = (cnts > num_resp)
plabels = np.concatenate((labels[cutoff], ['Other']))
pcnts = np.concatenate((cnts[cutoff], [cnts[~cutoff].sum()]))

# Plot
fig, ax = plt.subplots(figsize=(8, 8))
ax.pie(pcnts, labels=plabels, autopct='%1.1f%%')
ax.set_title("Country of Residence");
fig.tight_layout()

# Map countries to continents
import pycountry_convert as pc
cont_code_to_cont_name = {
    'NA': 'North America',
    'SA': 'South America',
    'AS': 'Asia',
    'EU': 'Europe',
    'AF': 'Africa',
    'OC': 'Oceania',
}
def country_to_continent(country_name):
    cc = pc.country_name_to_country_alpha2(country_name)
    cont_code = pc.country_alpha2_to_continent_code(cc)
    return cont_code_to_cont_name[cont_code]
c2c = np.vectorize(country_to_continent, otypes='U')

# Organize countries below the privacy cutoff by their continent
remaining_countries = labels[~cutoff]
continents = c2c(remaining_countries)
with open('_generated/countries_by_continent.md', 'w') as of:
    of.write('|  |  |\n')
    of.write('|---------------|-------------|\n')
    for continent in np.unique(continents):
        clist = remaining_countries[continents == continent]
        of.write(f"| **{continent}:** | {', '.join(clist)} |\n")

glue('2021_num_unique_countries', len(labels), display=False)
glue(
    '2021_num_country_respondents',
    gluval(country.shape[0], data.shape[0]),
    display=False
)


Africa:	Egypt, Ghana, Kenya, Nigeria, South Africa, Zimbabwe
Asia:	Bangladesh, China, Indonesia, Israel, Malaysia, Pakistan, Singapore, South Korea, Turkey
Europe:	Austria, Belgium, Bulgaria, Denmark, Finland, Hungary, Iceland, Ireland, Italy, Latvia, Norway, Poland, Russia, Slovakia, Spain, Sweden, Switzerland, Ukraine
North America:	Canada, Cuba, Mexico
Oceania:	Australia, Vanuatu
South America:	Argentina, Chile, Colombia, Ecuador, Paraguay, Peru, Venezuela

Education¶

440 (84%) respondents shared their education history, spanning the range from pre-highschool graduation through Doctorate level with many other specialist degrees.

Generally, respondents are highly educated. Nine out of ten have at least a Bachelor’s degree and one in three holds a PhD.

The following figure summarizes the distribution for the highest degrees obtained by respondents.

degree = data['degree'][data['degree'] != '']
labels, cnts = np.unique(degree, return_counts=True)

fig, ax = plt.subplots(figsize=(8, 8))
ax.pie(cnts, labels=labels, autopct='%1.1f%%', labeldistance=None)
ax.legend()
ax.set_title("Highest Level of Education");
fig.tight_layout()

glue('2021_num_education', gluval(degree.shape[0], data.shape[0]), display=False)

Job Roles¶

205 (47%) of the 435 respondents who shared their occupation identify as an Data scientist, Scientist (in academia), or Software engineer.

role = data['role'][data['role'] != '']
labels, cnts = np.unique(role, return_counts=True)

# Sort results by number of selections
inds = np.argsort(cnts)
labels, cnts = labels[inds], cnts[inds]

fig, ax = plt.subplots(figsize=(12, 8))
ax.barh(np.arange(len(cnts)), cnts, align='center')
ax.set_yticks(np.arange(len(cnts)))
ax.set_yticklabels(labels)
ax.set_xlabel("Number of Respondents")
ax.set_title("Current Role");
fig.tight_layout()

glue('2021_num_occupation', role.shape[0], display=False)
glue(
    '2021_num_top_3_categories',
    gluval(cnts[-3:].sum(), role.shape[0]),
    display=False,
)
glue('2021_top_3_categories', f"{labels[-3]}, {labels[-2]}, or {labels[-1]}", display=False)

2020 Community Survey Respondents¶

4 (1%) of respondents shared whether or not they responded to the 2020 Community Survey. Only 25.0 percent of people reported having completed last year’s survey.

# Ignore empty fields and "prefer not to answer"
drop = np.logical_and(data['surv2020'] != '', data['surv2020'] != 'Not sure')
surv2020 = data['surv2020'][drop]
labels, cnts = np.unique(surv2020, return_counts=True)

glue('2021_num_surv2020_respondents', gluval(surv2020.shape[0], data.shape[0]), display=False)
yes_percent = 100 * cnts[1].sum() / cnts.sum()
glue('2021_yes_percent', f"{yes_percent:1.1f}", display=False)

Experience and Usage¶

Programming Experience¶

68% of respondents have significant experience in programming, with veterans (10+ years) taking the lead. Interestingly, when it comes to using NumPy, noticeably more of our respondents identify as beginners than experienced users.

fig, axes = plt.subplots(1, 2, figsize=(12, 6))
# Ascending order for figure
ind = np.array([-1, 0, 2, 3, 1])
for exp_data, ax in zip(('programming_exp', 'numpy_exp'), axes):
    # Analysis
    prog_exp = data[exp_data][data[exp_data] != '']
    labels, cnts = np.unique(prog_exp, return_counts=True)
    cnts = 100 * cnts / cnts.sum()
    labels, cnts = labels[ind], cnts[ind]
    # Generate text on general programming experience
    glue(f'2021_{exp_data}_5plus_years', f"{cnts[-2:].sum():2.0f}%", display=False)
    # Plotting
    ax.bar(np.arange(len(cnts)), cnts)
    ax.set_xticks(np.arange(len(cnts)))
    ax.set_xticklabels(labels)
axes[0].set_title('General Programming Experience')
axes[0].set_ylabel('Percentage of Respondents')
axes[1].set_title('Experience with NumPy');
fig.autofmt_xdate();
fig.tight_layout();

Programming Languages¶

401 (77%) of survey participants shared their experience with other programming languages. 64% of respondents are familiar with C / C++, and 40% with Matlab.

pl = data['prog_lang'][data['prog_lang'] != '']
num_respondents = len(pl)
glue('2021_num_proglang_respondents', gluval(len(pl), data.shape[0]), display=False)

# Flatten & remove 'Other' write-in option
other = 'Other (please specify, using commas to separate individual entries)'
apl = []
for row in pl:
    if 'Other' in row:
        row = ','.join(row.split(',')[:-2])
        if len(row) < 1:
            continue
    apl.extend(row.split(','))
labels, cnts = np.unique(apl, return_counts=True)
cnts = 100 * cnts / num_respondents
I = np.argsort(cnts)
labels, cnts = labels[I], cnts[I]

fig, ax = plt.subplots(figsize=(12, 8))
ax.barh(np.arange(len(cnts)), cnts, align='center')
ax.set_yticks(np.arange(len(cnts)))
ax.set_yticklabels(labels)
ax.set_xlabel("Percentage of Respondents")
ax.set_title("Programming Language Familiarity")
fig.tight_layout()

# Highlight two most popular
glue('2021_num_top_lang', f"{cnts[-1]:2.0f}%", display=False)
glue('2021_top_lang', labels[-1], display=False)
glue('2021_num_2nd_lang', f"{cnts[-2]:2.0f}%", display=False)
glue('2021_second_lang', labels[-2], display=False)

84 (21%) percent of respondents reported familiarity with computer languages other than those listed above. Of these, Rust was the most popular with 17 (4%) respondents using this language. A listing of other reported languages can be found below (click to expand).

['"' '"cobol' '"tcl' '-' 'amigae' 'asembler mc68k' 'asembler x86'
 'assembly' 'awk' 'bash' 'basic' 'clojure' 'cython' 'dart' 'dpl' 'elixir'
 'elm' 'f#' 'forth' 'go' 'golang' 'haskel' 'haskell' 'idl' 'kotlin' 'lisp'
 'llvm ir' 'logo' 'lua' 'mathematica' 'matlab' 'mysql' 'nim' 'nim lang'
 'nix' 'none' 'none except python' 'objective-c' 'ocaml' 'pascal' 'perl'
 'php' 'prolog' 'python' 'rust' 'scala' 'scheme' 'shell' 'sql' 'stata'
 'swift' 'systemverilog' 't-sql' 'tcl' 'tcl/tk' 'tex' 'typescript' 'vba'
 'verilog' 'visual basic' 'visual foxpro' 'what does familiar mean?'
 'wolfram']

NumPy Community Survey Results - 2021

Contributions