4

I have implemented a python script, which imports Numpy and Pandas and I would like to run this script on Android. To be more precise, I would like to embed this script into an application.
I would like to know whether it is possible? If so, what are the best-practices to implement it?
I would greatly appreciate any help!

Bence
  • 155
  • 2
  • 3
  • 6

1 Answers1

7

If you do not want to build a website or app and have Python/Pandas running as a backend. You can use Kivy as a packager to run Python on Android. Further, if you check out the answer to this question it points to the documentation for using numpy too - which is to use a "recipe" for compilation.

If using Kivy and not using a pure python library - these recipes will need to be used or created if they do not exist. So with Pandas, you would need to build this recipe yourself. Even if you do build this resource, the size of trying to load Pandas (not to include the resources it can require when performing analysis on dataframes) might be a bottleneck if trying to include it directly in the app and it still might be better to do this in a backend situation.

Community
  • 1
  • 1
LinkBerest
  • 1,281
  • 3
  • 22
  • 30
  • Thank you very much for your reply! I've started to use kivy and managed to create the necessary recipes (pandas wasn't among these after all). Having these apk files, I would like to ask whether it is possible to use them in android development (i.e. import the numpy package). – Bence Sep 24 '16 at 11:53
  • pandas is big so using it in mobile (directly on the device) would be difficult no matter the package used - which is why most move to using numpy for calculations and then matplot directly for visualizations. If you have the apk -> [I would follow the instructions from kivy](https://kivy.org/docs/guide/packaging-android.html). After that it is a process so [ask direct questions](http://stackoverflow.com/help/how-to-ask) when you run into problems – LinkBerest Sep 24 '16 at 12:22