The ability to programmatically and reliably infer the social attributes of a person from their name can be useful for a broad set of tasks, from estimating bias in coverage of women in the media to estimating bias in lending against certain social groups. But unlike the American Census Bureau, which produces a list of last names and first names, which can (and are) used to infer the gender, race, ethnicity, etc., from names, the Indian government produces no such commensurate datasets. Hence inferring the relationship between gender, ethnicity, language group, etc., and names has generally been done with small datasets constructed in an ad-hoc manner.

We fill this yawning gap. Using data from the Indian Electoral Rolls (parsed data here), we estimate the proportion female, male, and [third sex]{.title-ref} (see here) for a particular [first name, year, and state.]{.title-ref}

Please also check out pranaam that uses land record data from Bihar to infer religion based on the name. The package uses indicate to transliterate Hindi to English.

Check out our interactive Streamlit App to test naampy with your own names!

• 🚀 Easy to use: Simple API with just two main functions
• 📊 Data-driven: Based on millions of names from Indian Electoral Rolls
• 🎯 Accurate: Provides confidence scores with predictions
• 🗺️ State-specific: Get region-specific predictions for better accuracy
• 🤖 ML-powered: Neural network fallback for names not in database
• 📈 Comprehensive: Covers 31 states and union territories

• Python 3.11
• pip or uv package manager

We strongly recommend installing naampy inside a Python virtual environment (see venv documentation):

pip install naampy

Or if you're using uv:

uv pip install naampy

To install the latest development version:

git clone https://github.com/appeler/naampy.git
cd naampy
pip install -e .

import pandas as pd
from naampy import in_rolls_fn_gender, predict_fn_gender

# Create a DataFrame with names
names_df = pd.DataFrame({'name': ['Priyanka', 'Rahul', 'Anjali']})

# Get gender predictions from electoral roll data
result = in_rolls_fn_gender(names_df, 'name')
print(result[['name', 'prop_female', 'prop_male']])

For names not in the electoral roll database:

# Use the neural network model for predictions
names = ['Aadhya', 'Reyansh', 'Kiara']
predictions = predict_fn_gender(names)
print(predictions)

import pandas as pd
from naampy import in_rolls_fn_gender

# Sample data
names = [{'name': 'gaurav'}, {'name': 'yasmin'}, {'name': 'deepti'}]
df = pd.DataFrame(names)

result = in_rolls_fn_gender(df, 'name')
print(result[['name', 'n_male', 'n_female', 'prop_female', 'prop_male']])

Output:

     name    n_male  n_female  prop_female  prop_male
0  gaurav   25625.0      47.0     0.001831   0.998169
1  yasmin      58.0    6079.0     0.990549   0.009451
2  deepti      35.0    5784.0     0.993985   0.006015

from naampy import predict_fn_gender

# Names not in electoral roll database
names = ["nabha", "hrithik", "kiara", "reyansh"]
predictions = predict_fn_gender(names)
print(predictions)

Output:

      name pred_gender  pred_prob
0    nabha      female   0.755028
1  hrithik        male   0.922181
2    kiara      female   0.614125
3  reyansh        male   0.891234

When you first run in_rolls_fn_gender, it downloads data from Harvard Dataverse to a local cache folder. Subsequent runs use the cached data for faster performance.

The package provides two complementary approaches:

1. Electoral Roll Data: Statistical data from millions of Indian voters
2. Machine Learning Model: Neural network trained on name patterns

For names not found in the electoral roll database, the package automatically falls back to the ML model.

For comprehensive documentation, examples, and API reference, visit: https://appeler.github.io/naampy/

Suriyan Laohaprapanon, Gaurav Sood, and Rajashekar Chintalapati

• appeler/pranaam — Predict religion based on names
• appeler/outkast — Map last names to caste categories
• appeler/parsernaam — AI-powered name parsing