The main results that users will be interested in can be seen by selecting the View Cultivar Coefficient button (Figure 10). This will open an editor with the final values of the estimated coefficients in it. The format of the file is the same as the CUL file for the selected crop, so one can copy this new set of cultivar coefficients into the appropriate CUL file to use in additional simulations. Note that one should use the DSSAT feature to Update all Lists after adding a new cultivar to any CUL file.
All of the outputs of DSSAT and GLUE running are saved in the C:\DSSAT45\GLWork\ directory. The contents of main output files are briefly described as follows:
(a) | Optimal Parameters. The optimal parameter set that was chosen through GLUE procedure was saved as a CUL file named according to the name and ID of the selected cultivar when generating the batch file. For example, if the selected cultivar was soybean COBB, then the CUL file is SBIB0002 COBB.CUL (Table 1 in Appendix E). |
(b) | Statistics of Posterior Distributions (Mean, Standard Deviation, and Maximum Likelihood Values). The two files identified as PosteriorDistribution_1.txt and PosteriorDistribution_2.txt (Table 2 and 3 in Appendix E) store the posterior distributions for each round of GLUE, including the mean values, standard deviations, and the parameter set that has the highest likelihood value in that round of GLUE. |
(c) | Empirical Distribution of Parameter Tables. The two files identified as RandomParameterSetsAndProbability_1.txt and RandomParameterSetsAnd |
Probability_2.txt (Table 4 and 5, Appendix E) store the really used parameter sets and their corresponding probability or normalized likelihood values for each round of GLUE.
(d) | Generated Parameter Sets. The two files identified as RealRandomSets_1.txt and RealRandomSets_2.txt store the really used parameter set in each round of GLUE. |
(e) | Last Model Run Results. Evaluate_output.txt stores the content of output file Evaluate.OUT of DSSAT for each model run. Since the Evaluate_output.txt is processed after each model run, only the result of last model run will be available in the Evaluate_output.txt file after the GLUE procedure. This file is not needed for result analysis, but it is described here because it will be in the directory and model users should ignore it. |
(f) | Results for Computing Likelihood Values. The two files EvaluateFrame_1.txt and EvaluateFrame_2.txt store the appended data of the processed Evaluate_output.txt files for the two rounds of GLUE. In each file, the simulated and measured outputs are saved for each treatment and each model run. |
(g) | Combined Likelihood Value for Each Parameter Set. The two files identified as IntegratedLikelihoodMatrix_Frame_1.txt, and IntegratedLikelihoodMatrix_Frame_2.txt (not shown) store the combined likelihood values for all treatments in each model run or for reach parameter set. For example, in the file IntegratedLikelihoodMatrix_Frame_1.txt, it stores the combined likelihood values for observations ADAP, MDAP, and PD1P for the first round GLUE. In IntegratedLikelihoodMatrix_Frame_2.txt, it stores the combined likelihood values for observations PWAM, HWAM, CWAM, LAIX, and L#SM for the second round GLUE. When the combined likelihood value is 1 in one column, it means the observation is absent. |
(h) | Combined Likelihood Value for Each Experiment Treatment. If there are only two treatments in the experiment for GLUE procedure, then the following files, IntegratedLikelihoodTreatment_1_1.txt, IntegratedLikelihoodTreatment_1_2.txt, IntegratedLikelihoodTreatment_2_1.txt, and IntegratedLikelihoodTreatment_2_2.txt, respectively, store the combined likelihood values for each treatment in each round of GLUE. The IntegratedLikelihoodTreatment_1_1.txt file, for example, stores the combined likelihood value for GLUE 1 and treatment 1 for all generated parameter sets, so do other files. One can see these files in the DSSAT45/GLWork directory after any GLUE estimation procedure is run. |
|