Stata Session 1

* ========================================
* SETUP - Run this at the start of every script
* ========================================

* Step 1: Clear everything from memory
* (This ensures you're starting fresh, not using old data)
clear all

* Step 2: Close any open log file
* (capture means "try this, but don't error if it fails")
capture log close

* Step 3: Don't pause after each screenful of output
set more off

* Step 4: Set your working directory
* IMPORTANT: Change this path to YOUR project folder!
* To find your path: On Mac, right-click your project folder in Finder
* → 'Get Info' → copy the path next to 'Where'.
* On Windows, open the folder in Explorer → click the address bar → copy.
* Use forward slashes (/) in Stata even on Windows.
cd "/Users/yourname/Dropbox/my_project"

* Step 5: Start a log file (records everything you do)
* "replace" overwrites the old log if it exists
log using my_analysis.log, replace

# ========================================
# SETUP - Run this at the start of every script
# ========================================

# Step 1: Clear environment (remove all objects)
rm(list = ls())

# Step 2: Set working directory
# IMPORTANT: Change this path to YOUR project folder!
setwd("/Users/yourname/Dropbox/my_project")

# Step 3: Load required packages
pacman::p_load(tidyverse, haven)

# ========================================
# SETUP - Run this at the start of every script
# ========================================

# Step 1: Import required packages
import pandas as pd
import numpy as np
import os

# Step 2: Set working directory
# IMPORTANT: Change this path to YOUR project folder!
os.chdir("/Users/yourname/Dropbox/my_project")

* Load the data
use mydata.dta, clear

* ----- EXPLORATION COMMANDS -----

* browse: Opens a spreadsheet view
* USE THIS FIRST to see what your data looks like
browse

* describe: Shows variable names, types, and labels
* This tells you WHAT variables you have
describe

* summarize: Shows mean, sd, min, max for numeric variables
* This tells you the RANGE of your data
summarize

* summarize specific variables
summarize age income

* summarize with detail: adds percentiles
* Useful for seeing the full distribution
summarize income, detail

* tabulate: Counts observations by category
* ESSENTIAL for categorical variables
tabulate sex

* See the numeric codes behind the labels
tabulate sex, nolabel

# Load the data
df <- read_dta("mydata.dta")

# ----- EXPLORATION COMMANDS -----

# View(): Opens a spreadsheet view
# USE THIS FIRST to see what your data looks like
View(df)

# glimpse(): Shows variable names and types
# This tells you WHAT variables you have
glimpse(df)

# summary(): Shows summary stats for all variables
summary(df)

# Summary for specific variables
summary(df$age)
summary(df$income)

# Detailed summary with percentiles
pacman::p_load(psych)
describe(df$income)

# table(): Counts observations by category
# ESSENTIAL for categorical variables
table(df$sex)

# With percentages
prop.table(table(df$sex))

# Load the data
df = pd.read_stata("mydata.dta")

# ----- EXPLORATION COMMANDS -----

# head(): Shows first few rows
# USE THIS FIRST to see what your data looks like
df.head(10)

# info(): Shows variable names and types
# This tells you WHAT variables you have
df.info()

# describe(): Shows summary stats for numeric variables
df.describe()

# Summary for specific variables
df[['age', 'income']].describe()

# Detailed summary with percentiles
df['income'].describe(percentiles=[.1, .25, .5, .75, .9])

# value_counts(): Counts observations by category
# ESSENTIAL for categorical variables
df['sex'].value_counts()

# With percentages
df['sex'].value_counts(normalize=True)

* ----- SELECTING OBSERVATIONS -----

* Keep only women (where sex == 2)
* "==" means "is equal to" (double equals for comparison)
keep if sex == 2

* Check how many observations remain
count

* Keep only ages 25-54
* "&" means AND - both conditions must be true
keep if age >= 25 & age <= 54

* Alternative: drop observations you don't want
* "|" means OR - either condition triggers the drop
drop if age < 25 | age > 54

* ----- SELECTING VARIABLES -----

* Drop variables you don't need
* This makes your dataset smaller and easier to work with
drop year serial pernum

* Keep only specific variables
keep id age income sex education

# ----- SELECTING OBSERVATIONS -----

# Keep only women (where sex == 2)
df <- df %>% filter(sex == 2)

# Check how many observations remain
nrow(df)

# Keep only ages 25-54
df <- df %>% filter(age >= 25 & age <= 54)

# Alternative: drop observations you don't want
df <- df %>% filter(!(age < 25 | age > 54))

# ----- SELECTING VARIABLES -----

# Drop variables you don't need
df <- df %>% select(-year, -serial, -pernum)

# Keep only specific variables
df <- df %>% select(id, age, income, sex, education)

# ----- SELECTING OBSERVATIONS -----

# Keep only women (where sex == 2)
df = df[df['sex'] == 2]

# Check how many observations remain
len(df)

# Keep only ages 25-54
df = df[(df['age'] >= 25) & (df['age'] <= 54)]

# Alternative: drop observations you don't want
df = df[~((df['age'] < 25) | (df['age'] > 54))]

# ----- SELECTING VARIABLES -----

# Drop variables you don't need
df = df.drop(columns=['year', 'serial', 'pernum'])

# Keep only specific variables
df = df[['id', 'age', 'income', 'sex', 'education']]

* ----- INDICATOR VARIABLES (0/1) -----

* Method 1: Two steps (WRONG - creates missing values!)
gen married = 1 if marst == 1
* Problem: married is MISSING (not 0) when marst != 1
tab married, missing  // You'll see missing values

* Method 2: Two steps (CORRECT)
gen married = 1 if marst == 1
replace married = 0 if marst != 1
* Now married is 0 or 1 for everyone

* Method 3: One step (BEST - cleaner code)
gen married = (marst == 1)
* The expression (marst == 1) evaluates to 1 if true, 0 if false

* ----- CATEGORICAL VARIABLES -----

* Group number of children: 0, 1, 2, 3, 4+
gen kids_cat = nchild if nchild <= 4
replace kids_cat = 4 if nchild > 4

* Create dummy variables from categorical
* This creates Kids_1, Kids_2, Kids_3, Kids_4, Kids_5
tab kids_cat, gen(Kids_)

* ----- TRANSFORMED VARIABLES -----

* Log of income (useful for skewed distributions)
* Add 1 to handle zeros (ln(0) is undefined)
gen ln_income = ln(income + 1)

* Income in thousands
gen income_k = income / 1000

# ----- INDICATOR VARIABLES (0/1) -----

# Method 1: Using ifelse (explicit)
df <- df %>% mutate(married = ifelse(marst == 1, 1, 0))

# Method 2: Using as.integer (cleaner)
df <- df %>% mutate(married = as.integer(marst == 1))

# ----- CATEGORICAL VARIABLES -----

# Group number of children: 0, 1, 2, 3, 4+
df <- df %>% mutate(
  kids_cat = case_when(
    nchild <= 4 ~ nchild,
    nchild > 4 ~ 4
  )
)

# Create dummy variables from categorical
df <- df %>% mutate(
  Kids_0 = as.integer(kids_cat == 0),
  Kids_1 = as.integer(kids_cat == 1),
  Kids_2 = as.integer(kids_cat == 2),
  Kids_3 = as.integer(kids_cat == 3),
  Kids_4 = as.integer(kids_cat == 4)
)

# ----- TRANSFORMED VARIABLES -----

# Log of income
df <- df %>% mutate(ln_income = log(income + 1))

# Income in thousands
df <- df %>% mutate(income_k = income / 1000)

# ----- INDICATOR VARIABLES (0/1) -----

# Simple boolean comparison (returns True/False, converted to 1/0)
df['married'] = (df['marst'] == 1).astype(int)

# ----- CATEGORICAL VARIABLES -----

# Group number of children: 0, 1, 2, 3, 4+
df['kids_cat'] = df['nchild'].clip(upper=4)

# Create dummy variables from categorical
dummies = pd.get_dummies(df['kids_cat'], prefix='Kids')
df = pd.concat([df, dummies], axis=1)

# ----- TRANSFORMED VARIABLES -----

# Log of income
df['ln_income'] = np.log(df['income'] + 1)

# Income in thousands
df['income_k'] = df['income'] / 1000

/*==============================================================================
    Master Do-File: My Research Project
    Author: Your Name
    Date: February 2026
==============================================================================*/

clear all
set more off

* Set the project root (CHANGE THIS!)
global root "/Users/yourname/Dropbox/my_project"

* Define paths
global build    "$root/build"
global analysis "$root/analysis"

* Run the build scripts
do "$build/code/01_clean_census.do"
do "$build/code/02_clean_policy.do"
do "$build/code/03_merge.do"

* Run the analysis
do "$analysis/code/01_summary_stats.do"
do "$analysis/code/02_main_regression.do"
do "$analysis/code/03_robustness.do"

#==============================================================================
#    Master Script: My Research Project
#    Author: Your Name
#    Date: February 2026
#==============================================================================

rm(list = ls())

# Set the project root (CHANGE THIS!)
root <- "/Users/yourname/Dropbox/my_project"

# Define paths
build <- file.path(root, "build")
analysis <- file.path(root, "analysis")

# Run the build scripts
source(file.path(build, "code", "01_clean_census.R"))
source(file.path(build, "code", "02_clean_policy.R"))
source(file.path(build, "code", "03_merge.R"))

# Run the analysis
source(file.path(analysis, "code", "01_summary_stats.R"))
source(file.path(analysis, "code", "02_main_regression.R"))
source(file.path(analysis, "code", "03_robustness.R"))

"""
Master Script: My Research Project
Author: Your Name
Date: February 2026
"""

import os
import subprocess

# Set the project root (CHANGE THIS!)
ROOT = "/Users/yourname/Dropbox/my_project"

# Define paths
BUILD = os.path.join(ROOT, "build")
ANALYSIS = os.path.join(ROOT, "analysis")

# Run the build scripts
exec(open(os.path.join(BUILD, "code", "01_clean_census.py")).read())
exec(open(os.path.join(BUILD, "code", "02_clean_policy.py")).read())
exec(open(os.path.join(BUILD, "code", "03_merge.py")).read())

# Run the analysis
exec(open(os.path.join(ANALYSIS, "code", "01_summary_stats.py")).read())
exec(open(os.path.join(ANALYSIS, "code", "02_main_regression.py")).read())
exec(open(os.path.join(ANALYSIS, "code", "03_robustness.py")).read())

* Import CSV with first row as variable names
import delimited "mydata.csv", varnames(1) clear

* Immediately check what you got
browse       // Look at the data
describe     // Check variable types
summarize    // Check value ranges

* Common issue: numbers imported as strings
* WHY? If your CSV has a comma in any number (like '1,234'), Stata reads
* the entire column as text. Same if there's any non-numeric character
* ('$50', 'N/A'). You'll know this happened if `summarize income` says
* 'no observations' or shows nothing.
* Fix by destringing (converting to numeric)
destring income, replace

# Import CSV (read_csv from tidyverse is best)
df <- read_csv("mydata.csv")

# Immediately check what you got
glimpse(df)    # Check variable types
summary(df)    # Check value ranges

# Note: read_csv automatically detects types
# It's smarter than base R's read.csv()

# Import CSV
df = pd.read_csv("mydata.csv")

# Immediately check what you got
df.head()     # Look at first rows
df.info()     # Check variable types
df.describe() # Check value ranges

# Note: pandas usually detects types correctly
# You can specify types if needed:
# df = pd.read_csv("mydata.csv", dtype={'id': str})

* Open Stata file
use "mydata.dta", clear

* Open specific variables only (saves memory for large files)
use age income education using "mydata.dta", clear

* Open only observations meeting a condition
use if age >= 25 using "mydata.dta", clear

# Need the haven package
pacman::p_load(haven)

# Read Stata file
df <- read_dta("mydata.dta")

# Note: haven preserves variable labels
# Access them with:
attr(df$income, "label")

# pandas can read Stata files directly
df = pd.read_stata("mydata.dta")

# Note: pandas preserves variable labels
# Access them with:
df.columns.to_list()  # Variable names

* Save as Stata file
save "mydata_clean.dta", replace

* Export as CSV
export delimited "mydata_clean.csv", replace

# Save as RDS (R's native format)
saveRDS(df, "mydata_clean.rds")

# Save as Stata file
write_dta(df, "mydata_clean.dta")

# Save as CSV
write_csv(df, "mydata_clean.csv")

# Save as pickle (Python's native format)
df.to_pickle("mydata_clean.pkl")

# Save as Stata file
df.to_stata("mydata_clean.dta")

# Save as CSV
df.to_csv("mydata_clean.csv", index=False)