2022-11-19-MOBS-GLEAM_FLUH.csv is missing forecasts for location 33 for every output type. These missing forecasts also have resulted in mislabeling the output type ids for all sample forecasts with a fips code greater than 33, as the indices for the 100 samples per location seem to be defined by 1:100 + (fips_ascending_index[i] -1). For example, location 01 (fips_ascending_index[1] = 1) has sample indices 1:100 because 1:100 + (1 -1) = 1:100.

This issue has carried over into hubExamples::forecast_outputs, which is generated using the data from that file.

Edit: To clarify, the problem presented above DOES NOT result in invalid samples in which the wrong samples are matched into a trajectory. It is more that the labeling scheme is not consistent across models (despite this being the original intention), and the labeling scheme for the MOBS model sample forecasts doesn't make logical sense in terms of human readability.