cea.analysis.sensitivity package¶

Submodules¶

cea.analysis.sensitivity.sensitivity_demand_analyze module¶

cea.analysis.sensitivity.sensitivity_demand_count module¶

Return the count a list of samples in a specified folder as input for the demand sensitivity analysis.

This reads in the samples.npy file produced by the script sensitivity_demand_samples.py and prints out the number of samples contained. This can be used for scripting the demand simulations with a load sharing facility system like the Euler cluster.

cea.analysis.sensitivity.sensitivity_demand_count.count_samples(samples_path)[source]¶

Read in the samples.npy numpy array from disk in the samples_path and report the row count (each row in the array is a sample to simulate for either the morris or the sobol method.

Parameters:	samples_path (str) – path to folder with the samples - see sensitivity_demand_samples.py
Returns:	number of samples in the samples folder.

cea.analysis.sensitivity.sensitivity_demand_samples module¶

cea.analysis.sensitivity.sensitivity_demand_simulate module¶

cea.analysis.sensitivity.sensitivity_optimization module¶

Note

documentation pending

Module contents¶

Sensitivity analysis for demand_main.py

These scripts use the morris algorithm (morris 1991)(campologo 2011) and Sobol Algorithm Sltalli 20110 to screen the most sensitive variables of a selection of parameters of the CEA.

The morris method serves to do basic screening of input variables and it is based on the family of One-at-a-time screening methods (OAT). morris provides a ranking but not a quantitative measure of the importance of each parameter.

The Sobol method serves for a complete sensitivity analysis of input variables. It is based on variance methods.