n_rows and n_columns must be positive integer. (Same as #853 and #836)

[WillTirone]

Describe the bug

This is similar to #853 and #836, but posting anyway just as another example with version info and a screenshot in Jupyter Notebook. Feel free to mark as a duplicate if desired.

When attempting the basic profile.to_widgets() example in the README, I encounter the attached error that n-rows and n-columns must be positive integers.

To Reproduce

follow the example in picture above

Version information:

Windows 10
Python: 3.8.5
Jupyter Notebook: 6.1.4
pandas-profiling: 3.1.0

Additional context

Just wanted to throw this out there since I've never used the profiling report and it seems the basic example is broken - I could be doing something obviously wrong though.

Happy to attempt a bug fix and PR if this is an actual issue. Let me know what else you all might need.

[sbrugman]

This is indeed a bug, a PR is very much welcome 👍

[WillTirone]

@sbrugman I will give it my best shot and let you know what I find!

[WillTirone]

@sbrugman this seems to be fixed in the current pandas-profiling version on GitHub (3.1.1). I've been playing around with it, but I just cloned the code and was trying to recreate examples when I noticed it works. Here's the example from #853 that now works but fails in 3.1.0.

Let me know if you need anything else to close this issue

[AdamXCA]

I have the same problem as this, and I fixed it by the steps below.

uninstall pandas profiling
update Anaconda
update Notebook
!pip install https://github.com/pandas-profiling/pandas-profiling/archive/refs/tags/v3.0.0.zip

let me know if it works for you

[haugstve]

@WillTirone
I don't think this is fixed in version 3.1.1

I installed pandas profiling using
pip install https://github.com/pandas-profiling/pandas-profiling/archive/master.zip

Relevant versions from conda list:

python                              3.10.1          h38b4d05_2_cpython    conda-forge
pandas                              1.3.5           py310hdd25497_0    conda-forge
pandas-profiling              3.1.1                    pypi_0    pypi 
ipykernel                          6.6.0           py310ha188af9_0    conda-forge
ipython                             7.30.1          py310h2ec42d9_0    conda-forge
ipython-genutils              0.2.0                    pypi_0    pypi
ipython_genutils              0.2.0                      py_1    conda-forge
ipywidgets                        7.6.5                    pypi_0    pypi
jupyter-client                    7.1.0                    pypi_0    pypi
jupyter_client                    7.1.0              pyhd8ed1ab_0    conda-forge
jupyter_core                      4.9.1           py310h2ec42d9_1    conda-forge
jupyterlab                          2.1.0                    pypi_0    pypi
jupyterlab-pygments        0.1.2                    pypi_0    pypi
jupyterlab-server              1.0.9                    pypi_0    pypi
jupyterlab-widgets           1.0.2                    pypi_0    pypi
jupyterlab_pygments       0.1.2              pyh9f0ad1d_0    conda-forge
jupyterlab_server             1.0.9                      py_0    conda-forge
jupyterlab_widgets           1.0.2              pyhd8ed1ab_0    conda-forge

I can't share my dataset so I'm not of much help. Maybe some of the others which have reported this #836 can be asked to test version 3.1.1?

[haugstve]

A quick follow up.

I just tried version 3.0.0 and it still didn't work on my dataset. The reason it doesn't work on my dataset might be more unrelated to the bug. Before I get the n-rows and n-columns must be positive integers I get a RuntimeWarning: overflow encountered in long_scalars. I like pandas profiling and appreciate the work you guys are putting in.

[prodramp]

I hit the same error with the titanic dataset.

!pip show pandas-profiling

Name: pandas-profiling
Version: 3.1.0
Summary: Generate profile report for pandas DataFrame
Home-page: https://github.com/pandas-profiling/pandas-profiling
Author: Simon Brugman
Author-email: [email protected]
License: MIT
Location: /Users/avkash/anaconda3/envs/python38/lib/python3.8/site-packages
Requires: htmlmin, jinja2, joblib, markupsafe, matplotlib, missingno, multimethod, numpy, pandas, phik, pydantic, PyYAML, requests, scipy, seaborn, tangled-up-in-unicode, tqdm, visions
Required-by:

Here is the titanic dataset url, I am using:

https://raw.githubusercontent.com/prodramp/publiccode/master/datasets/titanic.csv
Just use the following code to repro:

data_source_url = "https://raw.githubusercontent.com/prodramp/publiccode/master/datasets/titanic.csv"
df = pd.read_csv(data_source_url)
report = ProfileReport(df)
report.to_widgets()

[ApurvBhusari]

Instead of using to_widgets use to_notebook_iframe(). Hope it works for you