[PYTHON] iOS / iPad OS app "Juno" that allows machine learning on iPad

About Juno

https://apps.apple.com/jp/app/juno/id1462586500

You can download it at the app store for 1840 yen.

This app itself has been around for a long time, but since scikit-learn has finally become available in the recent app, I will purchase it and actually try it.

This app is basically an app that allows you to edit Jupyter Notebook on iOS / iPad OS.

Similarly, there is an app called carnets that can use Jupyter, and I used that because it is free, but scikit -learn could not be used.

Operation check

For the operation check, the code of this article will be used as it is.

In conclusion, everything works.

The main screen looks like this. An image that is more sophisticated as an iPad app than the Jupyter screen on a PC.

By the way, the above-mentioned carnets has a screen configuration similar to Jupyter on a PC.

The theme can be selected from white background and black background.

Of course, sklearn has been imported.

Complementation also works.

The pandas table is also displayed neatly.

The official apple keyboard for iPad does not have an esc key, but you can also assign the esc key to'@' etc., and there are plenty of hotkeys.

Of course, I can draw graphs and classify them.

From iOS / iPadOS 13, you can open any notebook you can see in a feature file.

It seems that the shell command can no longer be executed (?).

Instead of not being able to use the shell command, pressing on the filename will bring up a menu that allows you to install the missing package from the Install Python Package. Scikit-learn is included in the standard, but seaborn was not in the standard.

The installation was successful and pairplot worked.

As you can see in the image, are there any major storage formats?

When you select the format, the usual save screen is displayed, and of course you can send it to mac etc. with AirDrop etc.

The above is the feeling of touching Juno lightly.

The iPad is easy to carry around, so it may come in handy in cases like studying machine learning outside.