ucd-cws/calvin

1-year run is not running

msdogan opened this issue · 2 comments

msdogan commented

@jdherman @mfefer @josue-medellin
Previously, 82-year run was fine but small subset runs (1-year or small network) were not running. I was thinking that problem would be boundary condition difference between S09I05 and PyVIN. Now Justin updated data in HOBBES database with data from PyVIN (ucd-cws/calvin-network-tools@198f581). But, small subsets and 1-year model are still not running. We need to investigate more. Maybe run in debug mode and look at where we are getting debug flows. (82-year run is fine and running) Error code is below.

LP HAS NO PRIMAL FEASIBLE SOLUTION
glp_simplex: unable to recover undefined or non-optimal solution
If you need actual output for non-optimal solution, use --nopresol
Time used:   2.7 secs
Memory used: 59.9 Mb (62827240 bytes)
Writing basic solution to 'C:\Users\msdogan\AppData\Local\Temp\tmp4gzolcir.glpk.raw'...
120468 lines were written
[   25.53] Processing results
    Number of solutions: 0
    Solver results file: results.json
[   25.53] Applying Pyomo postprocessing actions
[   25.53] Pyomo Finished
msdogan commented

these are debug flows of 1-year run (Oct 2002-Sep 2003). Amounts are very small that's good news. @jdherman should we find the constraint and change a little bit so that it won't cause problem? And also we may need to do that for all years. Maybe Max's script would be useful here.

Link: DBUGSRC-C37, Time: 2003-08-31, Flow: 2.41610733905978e-16
Link: DBUGSRC-N201, Time: 2002-11-30, Flow: 0.000254582641544454
Link: DBUGSRC-SR_BUC, Time: 2003-09-30, Flow: 1.16151695450425e-05
Link: DBUGSRC-SR_GNT, Time: 2003-03-31, Flow: 0.000422804666095868
Link: DBUGSRC-SR_HID, Time: 2003-09-30, Flow: 0.00015654093552667
msdogan commented

Assuming new debug removal script would solve this issue.