loading the data

[dgcovell]

I am trying to work through your 'Clustering and Visualizing Documents using Word Embeddings', Readers and Williams. I appreciate making Jupyter note books available, but I am hoping that I can run directly in a python3 (version 3.6.8) on a linux system (Centos 7). I am able to import all the required programs (with the exception of ctfidf). However, I have not been able to load the data. Digging into the blogs, there appears to be a complication with docker. To summarize, the zipfile and zipurl designations do not give me access to 'clustering-visualizing-word-embeddings.zip', no do I have access to the zendo.org site. If possible, can you provide instructions to load the data.

Regards,
David

[jreades]

Hi David --

I'm having a bit of trouble making sense of your issue, can you confirm that you're not able to access this Zenodo link? If you can't access it then my only guess is that, for some reason, access to Zenodo is being blocked by your company/ISP or your company/ISP has been blocked by Zenodo. I can probably find a way to share this via another platform, but it's a strange issue to have.

On your second question about switching from a notebook to running a Python script directly, you can convert notebooks to Python using nbconvert; see this StackOverflow answer which should still work without issues (not at my computer at the moment).

Finally, you might find it easier to start from the full notebook that is available in my PH repo here.

Happy to try to resolve this in this repo thread, but do feel free to open up any other issues you encounter in the ph-tutorial-code repo.

Jon