Add category analysis for law enforcement facilities and misleading categories

[mao1756]

I added an analysis of which data points might be misleading and how to remove them. This addresses Issue #26. Summary:

Law enforcement facilities

It seems that wide variety of reports are concentrated at law enforcement facilities. Although we wouldn't check with TPD, it is possibly because they are where enforcement officers received the report. That is, they are not the location of the incident, rather where the incident report was received.

Following this hypothesis, we suggest not using data at LE facilities for pix2pix, but I believe using them for statistical analysis that does not involve the location of the crime is fine.

Misleading categories

"COMMUNITY POLICING", "ADMIN", "RECOVERED PROP", "AUTO THEFT RECOVERY", "UNABLE TO VERIFY" and "TRAFFIC" should not be included in our dataset.

• COMMUNITY POLICING: We already discussed this.
• ADMIN and TRAFFIC: There is no information on the TOPS website about these categories, so these reports are not informative.
• RECOVERED PROP and AUTO THEFT RECOVERY: These are not crime reports but about recovered properties.
• UNABLE TO VERIFY: These are for reports where officers could not find the incident.

I also have a code on how to remove these reports in the notebook.

[review-notebook-app[bot]]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[mao1756]

@ayhncgty Please review what I found, and if you agree, please approve the pull request. Once approved I will merge the PR.

[mao1756]

@abdullahnaeemmalik If you have time, please have a look on it as well.