quickeda-kit 0.1.0

A completely automatic EDA (Exploratory Data Analysis) library for Python

EDA Kit - Automatic Exploratory Data Analysis Library

A completely automatic Python library for performing comprehensive Exploratory Data Analysis (EDA) on any dataset. Just import, load your data, and run - the library handles everything from missing values to generating detailed reports.

Features

Fully Automatic: Just provide your dataset and the library does the rest
Intelligent Missing Value Handling: Automatically decides whether to use mean, median, mode, drop, or fill strategies
Comprehensive EDA: Statistics, distributions, correlations, outliers, and more
Visualizations: Automatic generation of plots and charts
Detailed Reports: Complete summary of all analysis performed
Interactive: Asks user questions when input is needed (train-test split, etc.)
Model-Ready: Prepares your dataset for machine learning model building

Installation

From PyPI (Recommended)

pip install quickeda-kit

From Source

git clone https://github.com/ashishali/auto_eda.git
cd auto_eda
pip install -e .

Dependencies

The package will automatically install all required dependencies:

• pandas>=1.3.0
• numpy>=1.21.0
• matplotlib>=3.4.0
• seaborn>=0.11.0
• scipy>=1.7.0
• scikit-learn>=1.0.0

Quick Start

Basic Usage

from eda_kit import AutoEDA
import pandas as pd

# Load your dataset
df = pd.read_csv('your_dataset.csv')

# Initialize AutoEDA
eda = AutoEDA(df=df)

# Run complete EDA workflow
results = eda.run_complete_eda()

From File Path

from eda_kit import AutoEDA

# Initialize with file path
eda = AutoEDA(file_path='your_dataset.csv')

# Run complete EDA workflow
results = eda.run_complete_eda()

Step-by-Step Usage

from eda_kit import AutoEDA
import pandas as pd

# Load dataset
df = pd.read_csv('your_dataset.csv')

# Initialize
eda = AutoEDA(df=df)

# Step 1: Handle missing values (automatic)
eda.handle_missing_values(auto=True)

# Step 2: Perform EDA analysis
eda.perform_eda(save_plots=True)

# Step 3: Generate report
eda.generate_report(save_to_file=True)

# Step 4: Get train-test split (interactive)
X_train, X_test, y_train, y_test = eda.get_train_test_split()

What Auto EDA Does

1. Missing Value Handling

• Automatic Detection: Identifies all missing values in the dataset
• Intelligent Strategy Selection:
  • Numeric columns: Uses mean for normal distributions, median for skewed data or outliers
  • Categorical columns: Uses mode (most frequent value)
  • High missing percentage (>50%): Drops the column
  • Time series: Uses forward fill
• User Control: Can be set to ask user for each column's strategy

2. Comprehensive EDA Analysis

• Basic Information: Dataset shape, memory usage, column types
• Descriptive Statistics: Mean, median, std, min, max for all numeric columns
• Distribution Analysis: Skewness, kurtosis, normality tests
• Outlier Detection: IQR and Z-score methods
• Correlation Analysis: Correlation matrix and highly correlated pairs
• Categorical Analysis: Value counts, unique values, most frequent categories

3. Visualizations

• Distribution plots for numeric columns
• Correlation heatmaps
• Box plots for outlier visualization
• Categorical value count charts

4. Report Generation

• Comprehensive text report with all findings
• Recommendations for next steps
• Summary of all transformations applied
• Saved to file (optional)

5. Train-Test Split

• Interactive prompts for target column selection
• Configurable test size
• Random state for reproducibility
• Returns ready-to-use train and test sets

Example Output

When you run run_complete_eda(), you'll see:

================================================================================
AUTOMATIC EDA WORKFLOW
================================================================================
Dataset shape: (1000, 10)
Columns: 10

Do you want to save EDA results to a directory? (yes/no) (default: yes): yes
Enter output directory path (press Enter for 'eda_output'): 
Results will be saved to: eda_output

================================================================================
STEP 1: HANDLING MISSING VALUES
================================================================================

Found missing values in 3 columns:
  - age: 50 missing (5.00%)
  - income: 20 missing (2.00%)
  - category: 10 missing (1.00%)

Automatically deciding strategies for handling missing values...

Strategies applied:
  - age: median
  - income: mean
  - category: mode

✓ Missing values handled successfully!

================================================================================
STEP 2: PERFORMING EXPLORATORY DATA ANALYSIS
================================================================================
Running comprehensive EDA analysis...
  - Gathering basic information...
  - Calculating descriptive statistics...
  - Analyzing distributions...
  - Detecting outliers...
  - Calculating correlations...
  - Analyzing categorical variables...
  - Generating visualizations...
EDA analysis complete!

================================================================================
STEP 3: GENERATING REPORT
================================================================================

================================================================================
AUTOMATIC EDA REPORT
================================================================================
...
[Detailed report with all findings]
...

================================================================================
EDA WORKFLOW COMPLETE!
================================================================================

Your dataset is now ready for model building!
Cleaned dataset shape: (1000, 10)
Original dataset shape: (1000, 10)

API Reference

AutoEDA Class

__init__(df=None, file_path=None)

Initialize AutoEDA with a dataframe or file path.

run_complete_eda(auto_handle_missing=True, save_plots=True, save_report=True, get_split=False, target_column=None)

Run the complete EDA workflow.

Parameters:

• auto_handle_missing (bool): Automatically handle missing values
• save_plots (bool): Save visualization plots
• save_report (bool): Save report to file
• get_split (bool): Perform train-test split
• target_column (str): Target column for split

Returns: Dictionary with cleaned dataframe, results, and split data

handle_missing_values(auto=True)

Handle missing values in the dataset.

perform_eda(save_plots=True)

Perform comprehensive EDA analysis.

generate_report(save_to_file=True)

Generate and save EDA report.

get_train_test_split(target_column=None, test_size=None, random_state=None)

Get train-test split with interactive prompts.

get_cleaned_dataframe()

Get the cleaned dataframe after EDA.

get_results()

Get all EDA results.

Requirements

• Python >= 3.7
• pandas >= 1.3.0
• numpy >= 1.21.0
• matplotlib >= 3.4.0
• seaborn >= 0.11.0
• scipy >= 1.7.0
• scikit-learn >= 1.0.0

Project Structure

datascience_package/
├── eda_kit/
│   ├── __init__.py
│   ├── auto_eda.py          # Main AutoEDA class
│   ├── missing_value_handler.py  # Missing value handling logic
│   ├── eda_analyzer.py       # EDA analysis functions
│   └── report_generator.py   # Report generation
├── setup.py
└── README.md

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

MIT License

Author

Ashish Y Beary