police work, exams, phone calls
Apr. 12th, 2006 01:47 pmso, some updates.
i've made 2 of my gazillion phone calls. both were painless. and enough of my flist convinced to feeling the same way about phone calls, that we could maybe start our own 12-step program or something...
the 2nd was to the police commander who e-mailed me about the consultant's job. so, some details:
the initial project was called "Neighborhood Safety Working Group" and was commissioned by the mayor when the homicide rate skyrocketed again a few years ago, to find out why, etc.
(side note: there was some significant 'gang' activity, although they weren't CALLING it that. they were called 'neighborhood disputes' in the media...)
anyway, they got a professor from a large college in the south of the state to try and work with the departments electronic police report system, collate that with court records, etc. a veritable SHITLOAD of data (and we all know my love of data, eh
divanoir?) Anyway, he hit a snag when trying to get the report data into something that was of use to most statistical software. apparently it's in an SQL relational data base (which may be redundant). each report can have many actual entries, especially since there would be separate ones for each offender, victim, property, etc. Up towards 500+ data fields. apparently, this professor couldn't figure out how to do it, or something. there may be more drama than that, which i'm not yet privileged enough to know :)
now, all these records are linked by the incident number. so in THEORY, one could create a data set with a hellaciously long record length, matching things up by incident number. i know how to do it, or at least what the logic would be. the question is if it can be done realistically. my hunch is it'd have to be done in chunks (maybe all january, all february, etc.). but i need to check with IT people before we make this a formal relationship.
so that's the deal thus far. :)
now i need to write an exam for my intro sociology class so i can give them a study guide (ugh). i hate writing study guides. and there's that boatload of phone calls to do, as well. i best get to it.
i've made 2 of my gazillion phone calls. both were painless. and enough of my flist convinced to feeling the same way about phone calls, that we could maybe start our own 12-step program or something...
the 2nd was to the police commander who e-mailed me about the consultant's job. so, some details:
the initial project was called "Neighborhood Safety Working Group" and was commissioned by the mayor when the homicide rate skyrocketed again a few years ago, to find out why, etc.
(side note: there was some significant 'gang' activity, although they weren't CALLING it that. they were called 'neighborhood disputes' in the media...)
anyway, they got a professor from a large college in the south of the state to try and work with the departments electronic police report system, collate that with court records, etc. a veritable SHITLOAD of data (and we all know my love of data, eh
![[livejournal.com profile]](https://www.dreamwidth.org/img/external/lj-userinfo.gif)
now, all these records are linked by the incident number. so in THEORY, one could create a data set with a hellaciously long record length, matching things up by incident number. i know how to do it, or at least what the logic would be. the question is if it can be done realistically. my hunch is it'd have to be done in chunks (maybe all january, all february, etc.). but i need to check with IT people before we make this a formal relationship.
so that's the deal thus far. :)
now i need to write an exam for my intro sociology class so i can give them a study guide (ugh). i hate writing study guides. and there's that boatload of phone calls to do, as well. i best get to it.