# 📌 Step 1: Import necessary libraries

import pandas as pd

import numpy as np

# 📌 Step 2: Load the dataset (Replace with your actual path or use an online URL if needed)

# Assuming your dataset is named '[Link]'

df = pd.read_csv("[Link]", index_col="Respondent") # 'Respondent' as unique identifier

# 📌 Step 3: Preview the data

print([Link](5)) # See the first 5 rows

print([Link]) # Show all column names

print([Link]()) # Get data types and non-null values

print([Link](include='all')) # Summary statistics (include non-numeric)

# 📌 Step 4: Check missing values

print([Link]().sum().sort_values(ascending=False)) # See which columns have most missing data

# 📌 Step 5: Drop columns with >80% missing values (optional, depends on your needs)

threshold = len(df) * 0.8

df = [Link][:, [Link]().sum() < threshold] # Keep only columns with <80% missing values

# 📌 Step 6: Fill missing values in specific columns

df['Student'] = df['Student'].fillna('Not a student') # Replace NaN in Student column

df['Age'] = df['Age'].fillna(df['Age'].median()) # Replace missing Age with median

# 📌 Step 7: Filter: Find all students who are unemployed

students_unemployed = df[(df['Student'].[Link]('Yes', na=False)) &

(df['Employment'] == 'Not employed, but looking for work')]

print(students_unemployed[['Country', 'Age', 'EdLevel']])

avg_age_per_country = [Link]('Country')['Age'].mean().sort_values(ascending=False)

print(avg_age_per_country)

# 📌 Step 9: Count how many people work with open source regularly

open_sourcers = df['OpenSourcer'].value_counts()

print(open_sourcers)

# 📌 Step 10: Use NumPy to analyze Age

ages = df['Age'].dropna().to_numpy()

print("Average Age:", [Link](ages))

print("Standard Deviation:", [Link](ages))

print("Minimum Age:", [Link](ages))

print("Maximum Age:", [Link](ages))

# 📌 Step 11: Crosstab - Gender vs Employment

gender_vs_employment = [Link](df['Gender'], df['Employment'], normalize='index')

print(gender_vs_employment)

# 📌 Step 12: Apply function to classify age groups

def age_group(age):

if age < 18:

return 'Teen'

elif age < 30:

return 'Young Adult'

elif age < 50:

return 'Adult'

else:
 return 'Senior'

df['AgeGroup'] = df['Age'].apply(age_group)

print(df[['Age', 'AgeGroup']].head())

# 📌 Step 13: Group by AgeGroup and count

agegroup_count = df['AgeGroup'].value_counts()

print(agegroup_count)

# 📌 Step 14: Which language was most commonly used?

# Many responses have multiple languages separated by semicolons, so we’ll split them and count

language_series = df['LanguageWorkedWith'].dropna().[Link](';')

all_languages = [Link]([[Link]() for sublist in language_series for lang in sublist])

print(all_languages.value_counts().head(10)) # Top 10 most popular languages

# 📌 Step 15: Save cleaned data for later use

df.to_csv("cleaned_stackoverflow.csv")

# 📌 Done!

# 📌 Step 1: Import the libraries

import pandas as pd

import numpy as np

# 📌 Step 2: Load the dataset

# 📌 Step 3: Clean and fill key identity columns

df['Gender'] = df['Gender'].fillna('Not disclosed')

df['Trans'] = df['Trans'].fillna('Not disclosed')

df['Sexuality'] = df['Sexuality'].fillna('Not disclosed')

df['Ethnicity'] = df['Ethnicity'].fillna('Not disclosed')

df['Age'] = df['Age'].fillna(df['Age'].median()) # Replace NaNs in Age with median age

# 📌 Step 4: View value counts

print("Gender Distribution:\n", df['Gender'].value_counts())

print("\nTrans Identity Distribution:\n", df['Trans'].value_counts())

print("\nSexuality Distribution:\n", df['Sexuality'].value_counts())

print("\nEthnicity Distribution:\n", df['Ethnicity'].value_counts().head(10)) # Top 10 ethnicities

# 📌 Step 5: Basic age statistics

print("\nAge Statistics:")

print("Mean:", df['Age'].mean())

print("Median:", df['Age'].median())

print("Min:", df['Age'].min())

print("Max:", df['Age'].max())

# 📌 Step 6: Filter people under 20 who identify as non-binary or gender diverse

young_nonbinary = df[(df['Age'] < 20) &

(~df['Gender'].[Link]('Man|Woman', na=False))]

print("\nYoung non-binary respondents:\n", young_nonbinary[['Country', 'Gender', 'Age',

# 📌 Step 7: Crosstab Gender vs. Employment

print("\nGender vs Employment:\n", gender_employment)

# 📌 Step 8: Crosstab Gender vs Career Satisfaction

gender_satisfaction = [Link](df['Gender'], df['CareerSat'], normalize='index')

print("\nGender vs Career Satisfaction:\n", gender_satisfaction)

# 📌 Step 9: Most common combinations of gender & sexuality

df['Gender_Sexuality'] = df['Gender'] + " | " + df['Sexuality']

print("\nTop Gender-Sexuality Combinations:\n", df['Gender_Sexuality'].value_counts().head(10))

# 📌 Step 10: Create age groups using NumPy

df['AgeGroup'] = [Link](df['Age'], bins=[0, 18, 25, 35, 50, 100],

labels=['<18', '18-25', '26-35', '36-50', '50+'])

print("\nAge group distribution:\n", df['AgeGroup'].value_counts())

# 📌 Step 11: Compare diversity across age groups

age_ethnicity = [Link](df['AgeGroup'], df['Ethnicity'], normalize='index')

print("\nEthnic distribution by AgeGroup:\n", age_ethnicity.head(5)) # First 5 age groups

# 📌 Step 12: Identify respondents who:

# - Are LGBTQ+ (not straight)

# - Have dependents

# - Are under 30

lgbtq_dependents_young = df[

(~df['Sexuality'].[Link]('Heterosexual|Straight', na=False)) &

(df['Dependents'] == 'Yes') &

(df['Age'] < 30)

]
print("\nLGBTQ+ respondents under 30 with dependents:\n",

lgbtq_dependents_young[['Country', 'Age', 'Sexuality', 'Dependents']].head())

# 📌 Step 13: Save results to CSV

df.to_csv("diversity_analysis_stackoverflow.csv")

# 📌 Step 1: Import libraries

import pandas as pd

import numpy as np

# 📌 Step 2: Load data

df = pd.read_csv("[Link]", index_col="Respondent")

# 📌 Step 3: Basic cleanup for relevant columns

# We'll work with: Compensation columns, Country, Employment, CareerSat, JobSat

df['CompTotal'] = pd.to_numeric(df['CompTotal'], errors='coerce')

df['CompFreq'] = df['CompFreq'].fillna('Yearly')

df['ConvertedComp'] = pd.to_numeric(df['ConvertedComp'], errors='coerce') # in USD

df['CareerSat'] = df['CareerSat'].fillna('No response')

df['JobSat'] = df['JobSat'].fillna('No response')

# 📌 Step 4: View compensation statistics

print(df['ConvertedComp'].describe())

# 📌 Step 5: Drop extremely high salaries (outliers > $300,000)

df_filtered = df[df['ConvertedComp'] < 300000]

# 📌 Step 6: Average compensation by country (Top 10)

avg_salary_by_country = df_filtered.groupby('Country')
['ConvertedComp'].mean().sort_values(ascending=False).head(10)

print("\nTop 10 countries by average compensation:")

print(avg_salary_by_country)

# 📌 Step 7: Median salary by job satisfaction level

salary_by_jobsat = df_filtered.groupby('JobSat')
['ConvertedComp'].median().sort_values(ascending=False)

print("\nMedian salary by job satisfaction level:")

print(salary_by_jobsat)

# 📌 Step 8: Count people earning over $100K by career satisfaction

rich_satisfied = df_filtered[df_filtered['ConvertedComp'] > 100000]

counts = rich_satisfied['CareerSat'].value_counts(normalize=True)

print("\nSatisfaction levels among those earning > $100K:")

print(counts)

# 📌 Step 9: Create income brackets

def income_bracket(salary):

if [Link](salary):

return 'Unknown'

elif salary < 20000:

elif salary < 50000:

return '20K–50K'

elif salary < 100000:

return '50K–100K'

elif salary < 150000:

return '100K–150K'

else:

return '150K+'

df_filtered['IncomeBracket'] = df_filtered['ConvertedComp'].apply(income_bracket)

print("\nIncome bracket distribution:")

print(df_filtered['IncomeBracket'].value_counts())

# 📌 Step 10: Satisfaction by income bracket (pivot)

satisfaction_income = [Link](df_filtered['IncomeBracket'], df_filtered['JobSat'], normalize='index')

print("\nJob satisfaction by income bracket:")

print(satisfaction_income)

# 📌 Step 11: Compare compensation frequency

comp_freq_counts = df['CompFreq'].value_counts()

print("\nCompensation frequency distribution:")

print(comp_freq_counts)

# 📌 Step 12: Median salary per compensation frequency

median_per_freq = [Link]('CompFreq')['ConvertedComp'].median()

print("\nMedian compensation by frequency:")

print(median_per_freq)
# 📌 Step 13: Save cleaned results

df_filtered.to_csv("compensation_analysis_stackoverflow.csv")