Pull Request for #1: Define structure and implement first version of code

[santiagohermo]

This is a review for #1, where we discussed and implemented a first version of the repository. Aggred-upon pipeline structure can be seen in https://github.com/JMSLab/LaroplanOCR/issues/1#issuecomment-1018890028. Can you review @jmshapir @miikapaal?

A couple of comments:

• You will see in the analysis/output/lgrYYYY_counts.csv files that there are several strings that are not words there. I drop many of those around here, but some still remain. We could go further in dropping those cases.
• Beyond the pipeline structure, I also updated the readme. It would be great if @jmshapir can take a look and implement improvements.

One final note. It might be useful to have someone new running the repo. We want to make sure that the requirements are complete and the explanation in the README is easy to understand.

[santiagohermo]

Thanks @miikapaal @jmshapir! I think all points, except the one about re-posting laroplaner (which will be tackled in #3), are resolved. I will thus make a clean run and wrap up this PR. After that I will open one more issue to come up with an illustration on how to use the data.

On the below point @jmshapir, which I forgot to reply before. Waiting sounds great! Before releasing the repo we can have someone run everything and make a final check.

Regarding asking someone to run the repo, I'm sure I can find someone to help when we're ready. It seems like it could be efficient to wait until we've completed this first pull, but if you see it differently, let me know!

[santiagohermo]

Summary here