I'm glad to see another take on the data. I'm also relieved to see that our results are along the same lines! :)
You say that "Red light violations" are automated. That didn't even cross my mind! I just assumed that they were manually issued without even thinking about it. I think it'd make more sense that it would be referring to automated red-light cameras, but are you basing that on anything or did you just make the opposite assumption?
I'd also just like to point out that I made an update yesterday with some more conventional statistics (standard deviation, Z-scores) and new weights accounting for days of the week.
Ah, I missed your update, thanks for posting that.
I guess we can't be totally sure that the red light tickets in question are automated. But it seems pretty likely given that Baltimore does have automated red light ticketing (http://cityservices.baltimorecity.gov/parkingfines/), and this dataset seemed to only contain automated versions of moving violations (ie, it had mobile and fixed speed cameras but not policeman-in-a-cruiser speeding tickets). Also it's just a ton of red light tickets, and in my personal experience street cleaning or meter tickets feel way more common than manually dispensed red light tickets.
I was a bit disappointed that this (and other datasets in Baltimore and SF) didn't have a codebook anywhere to explain the data. I'm guessing you weren't able to find one, either?
No, I wasn't able to find anything explaining the data either. I actually called a couple of offices to ask how complete it was, and nobody knew what to do with me. The first number I could find was the mayor's office. The person who answered transferred me to the DMV, and the lady who answered there had no idea what I was talking about and put me on hold. I hung up after a while. :/
You say that "Red light violations" are automated. That didn't even cross my mind! I just assumed that they were manually issued without even thinking about it. I think it'd make more sense that it would be referring to automated red-light cameras, but are you basing that on anything or did you just make the opposite assumption?
I'd also just like to point out that I made an update yesterday with some more conventional statistics (standard deviation, Z-scores) and new weights accounting for days of the week.
http://robert.io/posts/4.html#updates