CSVValidationTester

The CSVValidationTester can be used for validation purposes. It compares two CSV files: one produced by a simulation and one containing measured data. Then, it computes mean and standard deviation of relative or absolute error between measured and simulated data points.

Users can specify an upper bound for both the mean and std. deviation value via mean_limit and std_limit parameters, respectively.

Error type can be either 'relative' or 'absolute' and be set via err_type parameter.

Multiple files can be specified in cvsdiff parameter.

For inspecting the values, users can add -v flag when running the run_tests scripts and they can check the results that will come out in a table like this:

file                                     | computed             | requested
--------------------------------------------------------------------------------------
test_15.csv                              | 0.91 +/- 1.          | 1.00 +/- 2.00

Here file is the file name from csvdiff parameter, computed is the mean and std. deviation of the simulation values and requested is what was prescribed via mean_limit and std_limit

Note that currently only steady-state comparison is supported.