Historical Weather Charter

About

A historical weather charter web app that displays ~25 years of highs and mins on a per month scale for any input location.

Instructions to Run Locally:

1. Install Python 3.11.x
2. Create virtual environment

$ cd historical-weather-app
$ python3 -m venv env

3. Activate virtual environment and install packages

$ ./env/Scripts/activate  # For PowerShell
$ source env/bin/activate  # For ZSH
$ pip install --upgrade pip
$ pip install -r app/requirements.txt

4. Ensure that the model_backend in gbmodel/__init__ is set to sqlite3
5. Run app

$ python3 app/app.py

Instructions to Deploy on Cloud Run:

1. Clone repo in Cloud Shell
2. Ensure that the model_backend in gbmodel/__init__.py is set to datastore
3. Ensure that the CLIENT in gbmodel/model_datastore.py is set to the project name (same as ${GOOGLE_CLOUD_PROJECT})
4. Create docker container from historical-weather-charter/ directory
5. Create a service account called "weather" with the Cloud Datastore User role
6. Create Google Datastore database in "Datastore Mode" that resides in the US West1 region
7. Create a "weather" entity kind in datastore with a "[default]" namespace and "Numeric ID (auto-generated)" Key identifier. It should have the following properties:
   1. A "date" value
   2. A "lat_long" value
   3. A "max_temp" floating point number
   4. A "min_temp" floating point number
8. Create an index on the lat_long and date properties of the weather kind
9. Deploy to Cloud Run

$ git clone https://github.com/henrykaus/historical-weather-charter
# Make changes for steps 2 and 3
$ gcloud builds submit --timeout=900 --tag gcr.io/${GOOGLE_CLOUD_PROJECT}/gcp-weather
# Create and configure datastore db for steps 5-7
$ gcloud datastore indexes create index.yaml
$ gcloud run deploy gcp-weather --image gcr.io/${GOOGLE_CLOUD_PROJECT}/gcp-weather --service-account weather@${GOOGLE_CLOUD_PROJECT}.iam.gserviceaccount.com