##NOAA Weather Data Fun
This is a Django app with accompanying scripts that uses NOAA's QCLCD dataset to display graphs and other meaningful aggregations in a glitzy(ish) web frontend.
Some features:
- View a graph of high/low/avg temps for any of 2000 weather stations in the US, from 2011-present.
- View a seasonal summary of any weather station's data.
- Easily compare any two weather stations, see temperature charts overlaid and seasonal summaries side-by-side.
- Read a plain-English comparison of the two locations' annual conditions.
###If you're interested in playing around with this on your own, here are some not-so-simple instructions:
- Clone the repository
- Run manage.py syncdb to create your database schema. Note that privileged
settings.py
info is stored as environmental variables. - Run manage.py migrate --fake to run through all the old migrations. South will be very confused at this point and you'll probably need to manually update the schema / apply a migration.
- Delete all indexes and foreign key relations from your tables. Save the recreation statements, you'll need them later.
- Download the QCLCD datasets from NOAA's website using
scripts/noaa_urls.py
. Each month is about 80MB compressed, 400MB extracted. - Modify
scripts/noaa_set.py
with your folders, database passwords, etc. - Run
scripts/noaa_set.py
. Each year of data will require about 7GB in the database, and will take 4-6 hours to process and insert. - Look at the repository migrations and build your aggregation tables as needed.
- Recreate all the indexes that you previously deleted.
- Update the google maps API key in
templates/base.html
.