Closed jtwhite79 closed 2 years ago
No worries. I’ll go thru and make sure everything works again. Long weekend here so proly only get to it on tuesday
I found a little time today and was feeling guilty for breaking the part 1 notebooks so I patched those up so that the mechanics are at least happening now. But the posterior forecast unc in the k-only notebook is redic bad now for all forecasts so Im not sure if Ive goofed that up somehow...Im planning to work on a barebones ies and da notebook in the next couple days - historically this has led to revisiting the pstfrom setup process so just beware more mods maybe in store soon...
@rhugman Im sure I borked things with this - I renamed the head and sfr obs to avoid the underscore in the names (this helps later, trust me!) and I also switched the first stress period of the model from SS to transient (need this for sequential da later). Also reworked some of the pstfrom notebook to be explicit about setting obsvals (rather than just using the truth obsval csv files) since this is what happens in practice most of the time. Ive tried to patch the stuff I broke in the pstfrom notebook but this probably breaks some of the others - sorry man - I'll continue working thru these and patching. I just wanted to get this on your radar in case you are also working on these. I noticed the files in models/freyberg_pstfrom_pest_setup are not OS agnostic - on mac and linux, there is not "mult" subfolder, only files prepended with "mult/". Ive updated the models/freyberg_mf6_truth files and outputs to include these model changes (listed above) but couldnt run
hbd.make_truth()
so there are no longer ".meas" files, only the actual model output csv files (this is what Im pointing to in the pstfrom notebook now...).