search

CityofToronto / bdit_traffic_prophet

Suite of algorithms for predicting average daily traffic on Toronto streets
GNU General Public License v3.0
1 stars 1 forks source link

Rudimentary Connection Pipeline from Postgres to Traffic Prophet #22

Closed cczhu closed 4 years ago

cczhu commented 4 years ago

Basic capability to read raw data from our RDS Postgres DB. Because I don't know how to spin up a test Postgres instance, functions to read Postgres are not included in the test suite. It's therefore imperative for us to create a Jupyter Notebook of informal tests (which I suppose we can eventually stick in the sandbox after final merge), as well as perform a careful code review.

I've also changed some of the logic in the permanent count processing method in AnnualCount, partly to make it easier to read from Postgres (where, as shown in #15, we aggregate 15-minute count bins up to days in a Postgres materialized view before reading the view into Python) and partly to make the logic of AnnualCount more consistent than STTC_estimate3.m in TEPs-I. Now any day that doesn't have all 96 15-minute bins available is not used to estimate MADT.

Addresses #15.

cczhu commented 4 years ago

For all issues above, either resolved them or raised an issue. Ready to merge.

cczhu commented 4 years ago

The Jupyter Notebook of integration tests is now in the sandbox.