Maybe that is expected behaviour, but the sample dataset ist not converging?
Also the final cost after the second stage is far higher.
First stage:
Final Cost: error: 2.1598e+08 (mean res: 19.29, num: 562976), error valid: 2.1598e+08 (mean res: 19.29, num: 562976) FAILURE: Solver did not converge after maximum number of 50 iterations
Second stage:
Final Cost: error: 9.3924e+10 (mean res: 361.44, num: 562976), error valid: 9.3890e+10 (mean res: 361.50, num: 562624) NO_CONVERGENCE: Solver did not converge after maximum number of 50 iterations
Maybe that is expected behaviour, but the sample dataset ist not converging?
Also the final cost after the second stage is far higher.
First stage:
Final Cost: error: 2.1598e+08 (mean res: 19.29, num: 562976), error valid: 2.1598e+08 (mean res: 19.29, num: 562976) FAILURE: Solver did not converge after maximum number of 50 iterationsSecond stage:
Final Cost: error: 9.3924e+10 (mean res: 361.44, num: 562976), error valid: 9.3890e+10 (mean res: 361.50, num: 562624) NO_CONVERGENCE: Solver did not converge after maximum number of 50 iterations