Ammend Typos and Expand Mock-Data

[salvolpe]

I'm beginning to preprocess the data and I've noticed that there are a couple of one-letter typos in the labels:

• bluLipsFace --> blueLipsFace
• breathShortnes --> breathShortness
• gastrointestinnal --> gastrointestinal
• cadiosvascular --> cardiovascular
• comorbidites --> comorbidities

Also, I wanted to expand the data to include:

• Demographics, based on the options provided currently in Health
• • Specifically: race/ethnicity, age, gender, postal code
• Additional samples (maybe get up to at least 100 unique users, if not more)
• Eventually with more information on the effect of comorbidities, we can add illnesses to that category. But at the moment, I don't think we'll have any access to that information

[ngiangre]

Do we need to update the mock data based on the translations update @pavel-ilin?

[pavel-ilin]

Do we need to update the mock data based on the translations update @pavel-ilin?

Yes we do need to modify mock data. I can start working on that on Monday.

[ngiangre]

@salvolpe can you update on how usable the previous mock data was for your analysis? And recommend any useful modifications? Also update what analysis you did :)

[salvolpe]

@ngiangre @pavel-ilin The mock-data really wasn't useful, unfortunately, because of the limited relationships between the features and near-pure randomness of the construction. I'd love to set up a time to chat this weekend on ways that we make the mock data more robust for a model (that hopefully) works

[pavel-ilin]

I believe that issue is not ready to be closed. In pr #20 only show the structure of the mock data.

Next step will be to add logic of how we generate data and what parameter increase probability of having a covid.