pydatalens

pydatalens

pydatalens is a Python package designed to streamline the process of Exploratory Data Analysis (EDA), data cleaning, and visualization. It enables data scientists and analysts to quickly prepare, explore, and gain insights from datasets with minimal effort.

Features

1. Smart Summarization

• Automatically generates a summary of the dataset, including:
  • Data types
  • Missing values
  • Descriptive statistics
  • Unique value counts

2. Data Cleaning

• Detects and handles missing values using various strategies (mean, median, mode).
• Identifies and removes duplicate rows.
• Supports basic outlier detection (planned for future updates).

3. Correlation Analysis

• Generates a correlation matrix to identify relationships between features.
• Provides heatmaps for better visualization.

4. Automatic Visualizations

• Supports generating:
  • Histograms
  • Box plots
  • Correlation heatmaps
  • Scatter plots (planned for future updates).

5. Report Generation

• Exports EDA results and visualizations into a detailed HTML report for easy sharing.

Installation

Using pip (from source)

• Clone the repository:

  git clone https://github.com/gopalakrishnanarjun/pydatalens.git
  cd pydatalens
  

• Install the package:

  pip install -e .
  

Dependencies

• Python >= 3.6
• pandas >= 1.0
• numpy >= 1.18
• matplotlib >= 3.1
• seaborn >= 0.11

Install dependencies manually:

pip install pandas numpy matplotlib seaborn

Quick Start

1. Import the package

from pydatalens import eda, cleaning, visualizations

2. Load a dataset

import pandas as pd
df = pd.read_csv("your_dataset.csv")

3. Summarize the dataset

print(eda.summarize(df))

4. Handle missing values

df_cleaned = cleaning.handle_missing(df, strategy="mean")

5. Visualize the data

visualizations.plot_histogram(df_cleaned, column="age")
visualizations.correlation_heatmap(df_cleaned)

Examples

Summarizing the Data

from pydatalens import eda
summary = eda.summarize(df)
print(summary)

Cleaning the Data

from pydatalens import cleaning
df = cleaning.handle_missing(df, strategy="median")
df = cleaning.drop_duplicates(df)

Visualizing the Data

from pydatalens import visualizations
visualizations.plot_histogram(df, "column_name")
visualizations.correlation_heatmap(df)

Future Enhancements

• Advanced anomaly detection.
• Support for time series analysis.
• Enhanced visualization options (e.g., scatter plots, pair plots).
• Integration with machine learning pipelines.

Contributing

Contributions are welcome! If you'd like to contribute, please fork the repository and submit a pull request.

License

pydatalens is licensed under the MIT License. See the LICENSE file for more details.