View source: R/emulation_sensitivity_analysis.R
This method runs an ensemble for all parameter value sets specified in a CSV file generated from spartan latin-hypercube sampling techniques. The output is a CSV file that can then be analysed using the spartan analysis methods detailed in Technique 3.
1 2 3 4 5 |
filepath |
Path to where the analysis output should be stored |
surrogateModel |
Ensemble or emulator to use to emulate the simulation responses for those sets |
parameters |
Simulation parameter names |
measures |
Simulation output response names |
measure_scale |
Scale of each of the simulation responses |
param_file |
Name of the CSV file generated by spartan (or of) parameter values, separated in columns - if reading these in from a file |
dataset |
Name of the R dataset in the environment that contains the parameter sets (the tutorial one is emulated_lhc_values) |
ensemble_set |
Boolean stating whether this analysis is being run using an ensemble of machine learning methods (TRUE), or a single emulator (FALSE) |
normalise |
Whether the parameters in the sampling CSV file need to be normalised before input to the emulator (which must take values between 0 and 1) |
timepoint |
Simulation timepoint being analysed, if more than one. See the vignette for analysing more than one timepoint |
timepointscale |
Scale of the timepoints, if being used |
write_csv_files |
Whether output should be written to CSV file. Used with spartanDB |
graph_results |
Whether plots should be produced for this analysis |
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.