Running the backend regression tests yesterday and this morning, I picked up a test failure in forecast_calibration_test.py.
The test compares the forecast data on the ftp server for a specific base time to the forecast data stored in the database for the same base time and checks there isn't a divergence of more than 3%.
The test failure is specifically for Beijing, and for the pollutant o3
It would seem that the number from the FTP has been rounded and because the numbers are so small this is causing a massive differential causing the error.
Running the backend regression tests yesterday and this morning, I picked up a test failure in
forecast_calibration_test.py
.The test compares the forecast data on the ftp server for a specific base time to the forecast data stored in the database for the same base time and checks there isn't a divergence of more than 3%.
The test failure is specifically for Beijing, and for the pollutant o3