Dependency: Publish pyPRMS

[amsnyder]

pyPRMS is a library for working with PRMS-related files, include control/configuration files, climate forcings, data files, and outputs. pyPRMS includes the canonical metadata for PRMS that is defined in the documentation, but it has been coded into machine-readable xml files that can be utilized by the software.

Capabilities include:

• exploration of model setup files
  • get list of modules being used - param files often define more parameters than what is actually used for model to run; pyPRMS can help you reduce your parameters down to what is actually used by model
  • get parameter metadata that would be difficult to manually construct - Some parameters have adjustable information and are defined in input files with a variable. For example, the min/max bounds might be determined by number of segments. pyPRMS will construct these min/max values for you with information from the appropriate files that fills in those variable definitions.
  • identify missing parameters
  • check for degenerate parameters and expand them - these are parameters where the param file says dimensionality is 1, which means you should use the same value for all entries in dimension; this package can expand it out to full dimension
• model setup file editing
  • add missing parameters - adds with default values, then you can modify the values - step to write out to new parameter file
• read and write parameters in different file formats (ASCII, parameter database, netcdf), convert between formats - can create files for pywatershed or PRMS
• parameter file checks to look for errors such as:
  • for consistent parameter values
  • if all values are within min/max range
• read parameters into dataframes to be used in python workflows
• pull subsets of parameters
  • get parameter X for the following IDs
• geospatial/network exploration
  • get HRUs/segments upstream of POI
  • plot parameters by HRU or by segment
  • creates a DAG of stream network - can help you identify loops in network that shouldn't be there

This package definitely needs to be published for these notebooks to be released as it has not been documented/published in any format yet.

pyPRMS is a dependency of bandit.

Used in notebook 0b.

[amsnyder]

pywatershed package maintained by Development & Standard IWP Sub-program uses pyPRMS as a dependency, so hopefully they can be the long term product owners for this. I will reach out to Joe Hughes to discuss this.

[amsnyder]

We will plan to publish pyPRMS in EC github alongside pywatershed package. @pnorton-usgs will support the package's release and maintenance primarily under MAPPNAT time for now, but a long-term maintenance strategy still needs to be established. We may ask someone in the Development & Standards Sub-program to provide review for the software release and provide feedback to us on if they think the package needs to be run through the D&S Sub-program for refinement or if it should go straight to the Model Support Function. We would need to make a proposal in the Modeling Proposal Registry to get funding for either. In the meantime, we will plan for it to be a MAPPNAT product, though I haven't connected with Jacob on this specifically to verify, but @pnorton-usgs and I feel like this is the project that makes the most sense to maintain this software.

[rviger-usgs]

agree w/all that. to clarify, HyTEST will underwrite the publication but MAPPNAT will be the Product Owner of pyPRMS?

think dependency/risk re: establishment of long-term maintenance strategy should be called out in the FY25 HyTEST project plan. also, think this EC/HyTEST crossover or coordinated support should be called out as a "bright spot!"