quickeda-tool 0.1.0

Automated exploratory data analysis with a one-line API and a built-in Streamlit UI.

QuickEDA

Automated exploratory data analysis with a one-line API and a built-in Streamlit UI.

QuickEDA wraps every common EDA task — profiling, missingness, duplicates, outliers, statistics, correlation, time-series trends, target analysis, plain-language insights, and an exportable HTML report — behind one class and one command.

---

Install

pip install quickeda-tool

Three ways to use it

1. Launch the full Streamlit UI from Python

import quickeda
quickeda.launch()

That's it — your browser opens with the full UI: upload box, 8 analysis tabs, dynamic filters, interactive Plotly charts, and a one-click HTML report.

2. Pre-load a DataFrame into the UI

import pandas as pd
import quickeda

df = pd.read_csv("sales.csv")
quickeda.launch(df, name="sales")     # UI opens with data already loaded

3. Run analysis programmatically (notebooks)

from quickeda import AutoEDA
import pandas as pd

df = pd.read_csv("sales.csv")
eda = AutoEDA(df, name="sales")

eda.summary()              # plain-text summary in stdout
eda.show()                 # rich HTML view inline (Jupyter)
eda.to_html("report.html") # self-contained HTML report

# Filter and re-analyse
sub = eda.filter(region=["North", "South"], price=(100, 500))
sub.summary()

# Target-based EDA
result = eda.target("revenue")
print(result["correlations"])
print(result["insights"])

Or just use the command line

quickeda                         # launch UI, blank
quickeda data.csv                # launch UI with file pre-loaded
quickeda data.xlsx --port 8502   # custom port

What you get

Tab	What's inside
Overview	shape, memory, dtypes, preview, CSV download
Cleaning	missing values, duplicates, IQR outliers (analysis only)
Statistics	mean / median / std / skew, top-N for categoricals
Visualisations	Plotly histogram, boxplot, scatter, bar chart, heatmap
Time Series	auto datetime detect, rolling avg, peaks/troughs, seasonality
Target EDA	correlations or grouped comparisons, ranked importance
Insights	plain-language bullets — missingness, skew, dominance, outliers
Report	one-click self-contained HTML report

Public API

from quickeda import (
    AutoEDA,                     # one-class facade
    launch,                      # open the Streamlit UI

    # Direct functions (use any subset)
    load_dataset,                # CSV / Excel / JSON loader
    detect_column_types,
    missing_value_summary,
    duplicate_summary,
    detect_outliers_iqr,
    dataset_overview,
    numeric_statistics,
    categorical_statistics,
    generate_insights,
    correlations_with_target,
    group_means,
    categorical_importance,
    generate_target_insights,
    detect_datetime_columns,
    prepare_series,
    time_series_plot,
    trend_insights,
    build_html_report,
)

License

MIT