DataProfilerKit

A Python library that provides quick and insightful data profiling for pandas DataFrames. It generates detailed reports including missing values analysis, data type information, correlations, outliers, and column statistics in a clear, organized format.

Installation

pip install data-profiler-kit

Usage

from dataprofilerkit import DataProfiler
import pandas as pd

# Create or load your DataFrame
df = pd.read_csv('your_data.csv')

# Create a DataProfiler instance
profiler = DataProfiler(df)

# Generate the profile
profile = profiler.generate_profile()

# Access different aspects of the profile
print("Basic Information:")
print(profile['basic_info'])

print("\nMissing Values Analysis:")
print(profile['missing_values'])

print("\nColumn Statistics:")
print(profile['column_stats'])

print("\nDuplicates Analysis:")
print(profile['duplicates'])

print("\nOutliers Analysis:")
print(profile['outliers'])

Core Functionality

• Basic DataFrame Information:

  • Number of rows, columns, and total cells.
  • Memory usage of the DataFrame.
  • Data types and their counts.

• Missing Value Analysis:

  • Total missing values across the DataFrame.
  • Missing values by column.
  • Percentage of missing values for each column.

• Column-wise Analysis:

  • Numeric Columns:

    • Descriptive statistics (mean, median, standard deviation, etc.).
    • Skewness and kurtosis.

  • Categorical Columns:

    • Count of unique values.
    • Top 5 most frequent values with their percentages.

  • Datetime Columns:

    • Minimum and maximum values.
    • Range in days.

• Duplicate Detection:

  • Duplicate rows (count and percentage).
  • Duplicate columns (count and list of column names).

• Outlier Detection:

  • For numeric columns, detects outliers using:
    • Z-score method (with indices and percentages).
    • Interquartile Range (IQR) method (with indices and percentages).

Contributing

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

License

This project is licensed under the MIT License - see the LICENSE file for details.