Sweetviz

2.3.3 verified Mon Apr 27 auth: no python

A pandas-based library to visualize and compare datasets, providing an automated EDA report in HTML. Current version is 2.3.3, with releases about quarterly.

pip install sweetviz

Common errors

error ModuleNotFoundError: No module named 'sweetviz' ↓

cause Sweetviz is not installed in the current Python environment.

fix

Run pip install sweetviz.

error ImportError: cannot import name 'is_categorical_dtype' from 'pandas.api.types' ↓

cause In older sweetviz (<2.3.0) with newer pandas versions, the import path changed.

fix

Upgrade sweetviz to >=2.3.0.

error AttributeError: module 'numpy' has no attribute 'bool' ↓

cause Sweetviz versions <2.3.2 used deprecated numpy aliases removed in numpy 1.24+.

fix

Upgrade sweetviz to >=2.3.2.

error ValueError: The truth value of a DataFrame is ambiguous ↓

cause Passing a pandas DataFrame where a Series is expected, e.g., using analyze() with a multi-column target feature.

fix

Ensure target_feat is a single column name (string), not a list or DataFrame.

Warnings

breaking In v2.2.1, incompatibilities with Pandas 2.0+ and numpy 1.24+ were fixed. Upgrade from older versions (<2.2.1) may break if using newer pandas. ↓

fix Upgrade to sweetviz >=2.2.1.

deprecated The 'distutils' import was removed in v2.3.2; earlier versions may cause ImportError in Python 3.12+ where distutils is removed. ↓

fix Update sweetviz to >=2.3.2.

gotcha When using compare() or compare_intra(), the target feature must exist in both dataframes. Missing columns cause silent incorrect reports. ↓

fix Ensure all required columns are present in both datasets before comparison.

gotcha Large datasets (>100k rows) may cause high memory usage; the library loads the entire dataframe into memory and creates a large HTML report. ↓

fix Sample your data or use the 'verbosity' parameter (v2.3.0+) to limit detail. For massive datasets, consider alternatives like ydata-profiling.

Imports

sweetviz
wrong
```
from sweetviz import *
```
correct
```
import sweetviz
```
Module-wide import is standard; star imports are discouraged and may cause namespace pollution.

Quickstart

Generate an EDA report for the Iris dataset, targeting the 'species' column.

import pandas as pd
import sweetviz as sv

df = pd.read_csv('https://raw.githubusercontent.com/fbdesignpro/sweetviz/master/datasets/iris.csv')
report = sv.analyze(df, target_feat='species')
report.show_html('sweetviz_report.html')