pdcompare 0.1.6

Used to compare 2 Pandas DFs

pdcompare

Used to compare two pandas DataFrame objects to see how they changed.

pip install pdcompare

Requirements

• DataFrames must have the same index to compare correctly
• Warning will be thrown if the data types are non-numeric and you use .set_change_comparison(True)
• Warning will be thrown if the index names are different

STEPS

Initialize and call the compare() method:

from pdcompare import Compare

compare_object = Compare(df1,df2)
compare_object.compare()

To get a dictionary of the resulting comparison data call:

compare_object.output()

Output Details

Once you call the .output() method, you will receive a dictionary object in return. This dictionary has the following keys and associated values:

KEY	VALUE	VALUE Data Type
SUMMARY	high-level overview of differences	pd.DataFrame
ADDED	list of all index values that were added	pd.Series
ADDED_cols	list of all columns that were added	pd.Series
REMOVED	list of all index values that were removed	pd.Series
REMOVED_cols	list of all columns that were added	pd.Series
CHANGED	(see below for details)	pd.DataFrame

CHANGED output data

This data has the following columns

Column Header	Data
ID	Index value by which we tracked the alterations
COLUMN	Column that we saw an index change values
from	Value of specified column & index in the first table (old)
to	Value of specified column & index in the second table (new)

Examples

ScreamingFrog Crawl Comparison (SEO)

This is a great tool to compare crawls from different dates. Simply export the CSV files from ScreamingFrog. Then run this Google Colab notebook to create a Report in Google Sheets.

ScreamingFrog Crawl Compare in Colab

By default the code to connect to Google Sheets and do all the formatting is hidden, but feel free to peep behind the curtain to see how it was done. You can display the first block of code by opening using the drop-down triangle on the far left side of the block.

Weed Price Comparison

For a simple, get acquainted quickly example, use this. Thanks to Vicki for pointing me in the direction of these small datasets; and thanks to Frank BI for supplying the free datasets. I used Frank's weed price data from 2004 and compared them to 2005 across the 50 states. The example can be found in this repo's example folder.

Thanks for using my code

If you found this library useful, I'd appreciate a coffee.