I'm trying to validate my model in modelbuilder, but I don't exactly know which method is usefull for my situation. I've made a sutiability analysis based on certain criteria and I need to validate certain data sets of my model. I was told that computing a simple-square solution should be sufficient but I'm not certain on how to execute this.
Does anybody have experience with this?
More details on the inputs and desired outputs would help. What is the structure of your model (Screenshot)?
Model Builder > Run > Validate checks to ensure that the data exists, is valid for use with a tool, and that all tool parameter settings are correctly specified.
This is what 1 model looks like.
I want to do a suitability analysis on the best places to place hydrogen system in three scenarios, Solar / wind / Solar&wind, based on certain criteria; Solar irradiation, windspeeds mean, population density and slope. These criteria all have classifications ranging from 1-6 , meaning 1 is bad and 6 is really good. So for example, mean windspeeds from 1-3 m/s are category 1 and 11-15 m/s are category 6.
The desired output is a heatmap per scenario which shows areas above 500km2 with the best and worst combinations of the criteria, ranging from red(low score) to dark green (high score).
The inputs are:
The first one is solar data which is reclassified and then extracted for certain countries.
The second one is population denisty which is resampled, reclassified and extracted for the same countries as for the solar data.
The third one is the slope data for those same countries as for the others data sets.
These are all put together into a weighted sum and then a raster calculator has been performed to generate certain classes on the weight scale eg., 1-2 & 2-3 & 3-4 etc.
After that the data is crossed with a "restricted area" map and the polygons gotten from the final maps are then selected for only areas which are above 500km2.
Does this help?
then I am not sure what you mean by "validate". To me that means that it is useful in summarizing existing conditions and/or observations and you are trying to apply the model to other places based on the outcomes of the test area
This is what I mean with validation:
How to know if my implementation is correct?
In order to validate the function, I use a square with coordinates (0, 0), (0, 1), (1, 1), and (1,0) as input.
Because the input is simple, I am able to manually compute the area, which is 1.
Then I also output the area computed by the implemented function.
If the manually computed result and the result obtained by the function are the same,
then I know that the function works for the simple square at least;
otherwise, the implementation must be wrong.
But I don't know how to do this in ModelBuilder.
the only thing that comes remotely close is
which would require that it collects input values derived from an "extent" of some kind