Dany Chepenko

+1

The problem might occur due to the wrong input format of `item_features` when initializing `Dataset`. I would agree that the [documentation](https://github.com/lyst/lightfm/blob/1f3f35d2813393eb269073150fdf762ff19f0b42/lightfm/data.py#L197) is slightly misleading, it would be nice to see...

+1

@seratch I've encountered the same error. ``` object JsonLoader { def parseJsonConfig[T: Manifest](filename: String): T = { implicit lazy val formats: DefaultFormats = new DefaultFormats { override val strictOptionParsing: Boolean...

As it was mentioned above the compatibility issue caused this problem.

Should we close that issue?

Anyone working on it? Eager to contribute

@altern8tif, you can use my code sample. I haven't run it for a year or so; connecting to Xero was not an easy catch. ``` import json import requests import...

I might have missed something in official documentation, but I couldn't find a way to load the existed notebook. There is [only](https://nbformat.readthedocs.io/en/latest/api.html#nbformat.v4.new_notebook ) an option of creating a new notebook

Thanks, that is pretty much what I was looking for. For everyone who came across the same request I leave the updated code ``` nb = nbf.read('eda_new.ipynb', as_version=4) text =...