ValueError: ERROR for test model compatibility

[matthew....@gmail.com]

Hi all. I am running through the tutorials and am receiving this error when I run the  Test that Sound Mapping Tools is producing valid outputs. 

             File "C:\SMT\toolbox_V4_4_2\toolbox\scripts\validationhlpr.py", line 316, in sample_check

  raise ValueError("ERROR: %s value was not correctly computed: %s vs %s" % (check_name, test_value, test_values[count]))

ValueError: ERROR: eucdist_ft_raster value was not correctly computed: 4810 vs 9244

Failed to execute (validation2).

Any thoughts on what could be causing this? 

[matthew....@gmail.com]

For clarification, I'm running the test on the testing_v4_4 dataset where I copied the testing to my C:\smt\testing. Am I running the wrong data set?

[Dimitri Ponirakis]

I'm getting the same error

[Dimitri Ponirakis]

Does this mean my output raster is not matching?  How do I trace where the problem is?

[matthew....@gmail.com]

I think it could be a couple of things.

1) The cell size of two of the rasters are different; they all need to be the same.

2) The area boundary is interfering with propagation calculation. I found that if the area boundary was too close to the edge of any raster, an Euclidean Distance error message would be produced. I also received this error if the area boundary was small relative to the distance between/the number of points.

In response to your question about the data validation step, I think this is a result of a QA check performed near line 316 in the validationhlpr.py. I think if you change the test definition from =1 to =0 in the code the test will not run and you will not receive the error. 

I found that receiving this error does not mean that data are incompatible with the application. 

I also had problems with my data sets and received euclidean distance errors. I found that sometimes the datum, projection, or cell size became undefined when I cut or copied a raster. One way to test this is to use the Euclidean Distance tool in ArcToolbox and check if the distance and cell sizes are the same. If there is a value for the cell size when you input your raster and it is similar to other data sets, you likely won't get an error. If any of the inputs are blank after you add your raster, you will likely get a Euclidean Distance error.

Hope this helps.