Jupyter Notebook Pivottablejs Multiple Instances

Question

Intro

I am using a library that generates interactive pivot table/chart in Jupyter Notebook. This library was developed by @nicolaskruchten Also see documentations on Github

A Description Of The Problem

What I want to do is to call pivot_ui() multiple times within a single notebook page, by selecting "Run All" option under 'Cell'. Each time pivot_ui() is called, I pass different input data into it, and I am hoping to get different pivot charts/tables, each reflecting its corresponding input data.

test1.csv and test2.csv are 2 different data files. I want to create pivot charts on a same notebook page, subsequently. Here's my code:

# Cell 1
import pandas as pd
from pandas import DataFrame
from pivottablejs import pivot_ui

df1 = pd.read_csv("test1.csv")
pivot_ui(df)

# cell 2
df2 = pd.read_csv("test2.csv")
pivot_ui(df2)

If I run cell1 code only, then I get the pivot chart below.

If I run cell2 code only, then I get the pivot chart below.

If I run all cells, then the pivot chart for test1.csv is overwritten with test2.csv data, as shown below.

Is it possible to run all the cells, calling pivot_ui() multiple times with different input data each time, and display them accordingly in the notebook?

** Things I Have Tried**

1. I have looked through documentations on pivotUI, but I haven't found an input argument addresses this challenge.
2. I searched 'pivot_ui' in stackoverflow, and I can confirm no one else has asked the same questions.

I appreciate any suggestions/comments from the all mighty Internet community!

Answer 1

Solution: Set outfile_path keyword argument to a unique value for each pivot plot.

In the following example, test1.csv will generate pivottable_test1.html; test2.csv will generate pivottable_test2.html. Both html files are saved to disk. Jupyter notebook will show 2 different plots.

import pandas as pd
from pandas import DataFrame
from pivottablejs import pivot_ui
from IPython.display import HTML
df1 = pd.read_csv("test1.csv",)
pivot_ui(df1, outfile_path='pivottable_test1.html')

df2 = pd.read_csv("test2.csv",)
pivot_ui(df2, outfile_path='pivottable_test2.html')

Special thanks to Manusri Viswanathan for this solution.

Also see the solution provided by the author @nicolaskruchten via this link.