s3-datakit 0.3.8

A Python toolkit to simplify common operations between S3 and Pandas.

S3 DataKit 🧰

A Python toolkit to simplify common operations between Amazon S3 and Pandas DataFrames.

Key Features

• List files in an S3 bucket.
• Upload local files to S3.
• Download files from S3 directly to a local path or a Pandas DataFrame.
• Supports CSV and Stata (.dta) when reading into DataFrames.

Installation

pip install s3-datakit

or

uv add s3-datakit

Credential Configuration

This package uses boto3 to interact with AWS. boto3 will automatically search for credentials in the following order:

1. Environment variables (AWS_ACCESS_KEY_ID, AWS_SECRET_ACCESS_KEY, etc.).
2. The AWS CLI credentials file (~/.aws/credentials).
3. IAM roles (if running on an EC2 instance or ECS container).

For local development, the easiest method is to use a .env file.

1. Install python-dotenv in your project (not as a library dependency):

pip install python-dotenv

2. Create a .env file in your project's root:

AWS_ACCESS_KEY_ID=YOUR_ACCESS_KEY
AWS_SECRET_ACCESS_KEY=YOUR_SECRET_KEY
AWS_DEFAULT_REGION=your-region # e.g., us-east-1

3. Load the variables in your script before using s3datakit:

from dotenv import load_dotenv
import s3datakit as s3dk

# Load environment variables from .env
load_dotenv()

# Now you can use the package's functions
s3dk.list_s3_files(bucket="my-bucket")

Usage

List Files

import s3datakit as s3dk

file_list = s3dk.list_s3_files(bucket="my-data-bucket")
if file_list:
    print(file_list)

Upload a File

You can specify the full destination path in S3. If s3_path is not provided, the original filename from local_path is used as the S3 object key.

import s3datakit as s3dk

# Upload with a specific S3 path
s3dk.upload_s3_file(
    local_path="reports/report.csv",
    bucket="my-data-bucket",
    s3_path="final-reports/report_2025.csv"
)

# Upload using the local filename as the S3 key
# This will upload 'reports/report.csv' to 's3://my-data-bucket/report.csv'
s3dk.upload_s3_file(
    local_path="reports/report.csv",
    bucket="my-data-bucket"
)

Download a File

The download_s3_file function is versatile. You can download a file to a local path or load it directly into a Pandas DataFrame.

The download_s3_file function accepts the following parameters:

bucket (str): Required. The name of the S3 bucket where the file is located.

s3_path (str): Required. The full path (key) of the file within the bucket. local_path (str, optional): The local path where the file will be saved. If you don't provide this, the file will be saved in a data/ directory in your current working folder, using its original S3 filename.

to_df (bool, optional, default: False): If set to True, the function will attempt to read the downloaded file into a Pandas DataFrame. This is useful for .csv and Stata .dta files.

replace (bool, optional, default: False): If True, it will overwrite a local file if it already exists. By default, it skips the download if the file is already present to save time and bandwidth.

low_memory (bool, optional, default: True): When reading a CSV into a DataFrame (to_df=True), this is passed to pandas.read_csv to process the file in chunks, which can reduce memory usage for large files.

sep (str, optional, default: ","):**` The separator or delimiter to use when reading a CSV file into a DataFrame. For example, use '\t' for tab-separated files.

Option 1: Download to a local path

By default, if local_path is not provided, files are saved to a data/ directory in the current working directory.

import s3datakit as s3dk

# Download to a specific path
local_file = s3dk.download_s3_file(
    bucket="my-data-bucket",
    s3_path="final-reports/report_2025.csv",
    local_path="downloads/report.csv"
)
print(f"File downloaded to: {local_file}")

# Download to the default 'data/' directory, overwriting if it exists
s3dk.download_s3_file(
    bucket="my-data-bucket",
    s3_path="final-reports/report_2025.csv",
    replace=True
)

Option 2: Download directly to a Pandas DataFrame

import s3datakit as s3dk

df = s3dk.download_s3_file(
    bucket="my-data-bucket",
    s3_path="stata-data/survey.dta",
    to_df=True
)
print(df.head())