Pseudo data

Apart from the real data, the AmP procedure makes use of pseudo-data. Pseudo-data is a set of fixed (functions of) parameter values for a generalized animal species that is treated as (zero-variate) dependent data but with small weight coefficients. The mydata-file has the line [data, units, label, weights] = addpseudodata(data, units, label, weights);. This line adds pseudo-data to the data and (small) weight coefficients to the weights. The weights can be overwritten and set to zero, for instance, if you want to avoid its use in the estimation. The default weight coefficients for all real data sets (small or large) is 1, but that for a pseudo-data set is 0.1, except for kap_G, which is 20, and for k, which is 0, see addpseudodata.

The motivation for the use of pseudo data is to make sure that all key parameters are determined by data, but in a way that, if real data is already doing this, pseudo data hardly effect the resulting parameter values. The second motivation is to reduce the risk of finding a good fit for the wrong reasons. All parameters have a clear biological interpretation that must make sense. A good fit is not the only estimation criterion.

The following table presents pseudo-data values used by default in the AmP procedure. In the mydata file, current values can be modified by the user, weight coefficients can be set to zero if the user chooses to not include pseudo-data in the estimation and additional pseudo-data can be considered.

Code	Value	Unit	Description
v	0.02	cm/d	energy conductance
p_M	18	J/d.cm^3	volume-specific somatic maintenance costs
k	0.3	-	maintenance ratio
kap	0.8	-	allocation fraction to soma
kap_G	0.8	-	growth efficiency
kap_R	0.95	-	reproduction efficiency