Open dtordrup opened 5 years ago
I had a thorough check, and pretty much all of the parameter values are out of place. Below, I indicate after each row which parameter the value belongs to:
.par_value_eval
InterventionNoFGM, AnnualInflow = 0.5 * getAnnualInflowArray(modelTimeHorizon, countryISO3, baselineYear, secondarySexRatioFemale) 0.0818120054 <- FER
Intervention50, AnnualInflow = 0.5 * getAnnualInflowArray(modelTimeHorizon, countryISO3, baselineYear, secondarySexRatioFemale) 181323.0000000000 <- AnnualInflow
BAU, AnnualInflow = 0.5 * getAnnualInflowArray(modelTimeHorizon, countryISO3, baselineYear, secondarySexRatioFemale) 0.0250000000 <- SDE
InterventionNoFGM, AnnualInflow = 1.5 * getAnnualInflowArray(modelTimeHorizon, countryISO3, baselineYear, secondarySexRatioFemale) 0.2454360162 <- FER
Intervention50, AnnualInflow = 1.5 * getAnnualInflowArray(modelTimeHorizon, countryISO3, baselineYear, secondarySexRatioFemale) 543969.0000000000 <- AnnualInflow
BAU, AnnualInflow = 1.5 * getAnnualInflowArray(modelTimeHorizon, countryISO3, baselineYear, secondarySexRatioFemale) 0.4000000000 <- SDE
InterventionNoFGM, FER = 0.5 * getAnnualFertilityTransitionArray(modelTimeHorizon, baselineYear, countryISO3) 0.0002268884 <- GSA
Intervention50, FER = 0.5 * getAnnualFertilityTransitionArray(modelTimeHorizon, baselineYear, countryISO3) 0.0028719702 <- MRE
BAU, FER = 0.5 * getAnnualFertilityTransitionArray(modelTimeHorizon, baselineYear, countryISO3) 0.0001661492 <- MRA
InterventionNoFGM, FER = 1.5 * getAnnualFertilityTransitionArray(modelTimeHorizon, baselineYear, countryISO3) 0.0036302144 <- GSA
Intervention50, FER = 1.5 * getAnnualFertilityTransitionArray(modelTimeHorizon, baselineYear, countryISO3) 0.0459515236 <- MRE
BAU, FER = 1.5 * getAnnualFertilityTransitionArray(modelTimeHorizon, baselineYear, countryISO3) 0.0026583879 <- MRA
InterventionNoFGM, GMA = 0.25 * getModerateFGMAnnualRisk(incidenceFGM, countryISO3) 0.0001661492 <- MRE
Intervention50, GMA = 0.25 * getModerateFGMAnnualRisk(incidenceFGM, countryISO3) 0.0818120054 <- FER
BAU, GMA = 0.25 * getModerateFGMAnnualRisk(incidenceFGM, countryISO3) 181323.0000000000 <- AnnualInflow
InterventionNoFGM, GMA = 4 * getModerateFGMAnnualRisk(incidenceFGM, countryISO3) 0.0026583879 <- MRA
Intervention50, GMA = 4 * getModerateFGMAnnualRisk(incidenceFGM, countryISO3) 0.2454360162 <- FER
BAU, GMA = 4 * getModerateFGMAnnualRisk(incidenceFGM, countryISO3) 543969.0000000000 <- AnnualInflow
InterventionNoFGM, GSA = 0.25 * getSevereFGMAnnualRisk(incidenceFGM, countryISO3) 0.0250000000 <- SDE
Intervention50, GSA = 0.25 * getSevereFGMAnnualRisk(incidenceFGM, countryISO3) 0.0174704068 <- GMA
BAU, GSA = 0.25 * getSevereFGMAnnualRisk(incidenceFGM, countryISO3) 0.0002268884 <- GSA
InterventionNoFGM, GSA = 4 * getSevereFGMAnnualRisk(incidenceFGM, countryISO3) 0.4000000000 <- SDE
Intervention50, GSA = 4 * getSevereFGMAnnualRisk(incidenceFGM, countryISO3) 0.2795265096 <- GMA
BAU, GSA = 4 * getSevereFGMAnnualRisk(incidenceFGM, countryISO3) 0.0036302144 <- GSA
InterventionNoFGM, MRA = 0.25 * getMortalityRateAdults(countryISO3) 181323.0000000000 <- AnnualInflow
Intervention50, MRA = 0.25 * getMortalityRateAdults(countryISO3) 0.0250000000 <- DSE
BAU, MRA = 0.25 * getMortalityRateAdults(countryISO3) 0.0174704068 <- GMA
InterventionNoFGM, MRA = 4 * getMortalityRateAdults(countryISO3) 543969.0000000000 <- AnnualInflow
Intervention50, MRA = 4 * getMortalityRateAdults(countryISO3) 0.4000000000 <- DSE
BAU, MRA = 4 * getMortalityRateAdults(countryISO3) 0.2795265096 <- GMA
InterventionNoFGM, MRE = 0.25 * getMortalityRateElderly(countryISO3) 0.0028719702 <- MRE
Intervention50, MRE = 0.25 * getMortalityRateElderly(countryISO3) 0.0001661492 <- MRE
BAU, MRE = 0.25 * getMortalityRateElderly(countryISO3) 0.0818120054 <- FER
InterventionNoFGM, MRE = 4 * getMortalityRateElderly(countryISO3) 0.0459515236 <- MRE
Intervention50, MRE = 4 * getMortalityRateElderly(countryISO3) 0.0026583879 <- MRA
BAU, MRE = 4 * getMortalityRateElderly(countryISO3) 0.2454360162 <- FER
InterventionNoFGM, SDE = 0.25 * getType3DeinfibulationRate(countryISO3) 0.0174704068 <- GMA
Intervention50, SDE = 0.25 * getType3DeinfibulationRate(countryISO3) 0.0002268884 <- GSA
BAU, SDE = 0.25 * getType3DeinfibulationRate(countryISO3) 0.0028719702 <- MRE
InterventionNoFGM, SDE = 4 * getType3DeinfibulationRate(countryISO3) 0.2795265096 <- GMA
Intervention50, SDE = 4 * getType3DeinfibulationRate(countryISO3) 0.0036302144 <- GSA
BAU, SDE = 4 * getType3DeinfibulationRate(countryISO3) 0.0459515236 <- MRE
It would be useful to know, for a work-around, whether these values are only used for the labels on the tornado charts, or if they are also used as part of the plots or the underlying calculations?
This one is relatively complex to explain. I have a HEEMOD model with three strategies. I am running DSA with eight parameters, and ask the plot to evaluate parameter values.
The plot produces three tornado charts. The first one shows values for one parameter that this parameter never has:
As shown in the table below, GMA only ever has values <1.0:
A similar thing happens for plot #3:
And as can be seen, the MRA parameter never assumes values other than <1.0:
I therefore inspected the printed result of the DSA, and noticed the following (** added to highlight parameter names and (wrong) values):
So somehow the evaluated values for some of my parameters are being mixed up. All of the 181323 and 543969 values should be associated with the "AnnualInflow = .." parameters, i.e. first six rows in the printout just above. I haven't checked whether all other parameters are correct, these examples jump out because they are large numbers, but a few of the I can immediately tell are not right, e.g:
<- The value of GMA is constant, so of course 0.25*getModerate... and 4*getModerate... should be the same values.
Possibly importantly, many of the parameters depend on model time, and have separate values for each cycle (see the parameter table above). I have not used the model_time variable to create these parameters with simple arithmetic (as in many heemod examples), but defined these parameters using arrays of type numeric(..) and size = model time horizon. The parameter values shown for "AnnualInflow = ..." are the evaluated values for the fist model year (0.5 * 362,646).
Of course, it may be just a simple ordering issue...