AutoViML / AutoViz

Automatically Visualize any dataset, any size with a single line of code. Created by Ram Seshadri. Collaborators Welcome. Permission Granted upon Request.
Apache License 2.0
1.7k stars 196 forks source link

Failed DataFrames as input to the AutoViz #79

Closed albertoecf closed 1 year ago

albertoecf commented 1 year ago

Hey :D hope you're doing great. Thanks and congrats for all of your achievements

Feature:

Improved handling of pandas DataFrames as input to the AutoViz() function

Expected change in behavior:

The AutoViz() function should be able to handle pandas DataFrames as input, regardless of the data types of the columns or the presence of missing values.

How to test it:

Test the feature by passing a pandas DataFrame as input to the AutoViz() function and check that the visualizations are generated correctly.

Please note that the current version of autoviz library has a limitation when handling dataframes with missing values and columns with different data types.

Example code :

from autoviz.AutoViz_Class import AutoViz_Class

# Initialize the AutoViz class
AV = AutoViz_Class()

# Run the 'AutoViz' function
AV.AutoViz(df)

With df a pandas DataFrame with numeric non value column

Captura de pantalla 2023-01-17 a las 12 14 53

AutoViML commented 1 year ago

@albertoecf thanks for trying out AutoViz 👍

You cannot run AV.AutoViz() directly by sending in a dataframe. It assumes that the default for AV.AutoViz () is a filename.

Hence you need to send in the input like this:

# Run the 'AutoViz' function
AV.AutoViz(filename="". dfte=df)

Then it will work. That's how the syntax was designed, I may try to change it one day like the way you are using it. But no time.

Thanks Auto Vimal

albertoecf commented 1 year ago

Thanksss!! Properly working with the provided syntax

AV.AutoViz(filename="", dfte=df)