PAWN

PAWN is a global sensitivity analysis (GSA) algorithm ¹. Unlike variance based GSA approaches, it considers the entire distribution (specifically the CDF) of the model output. I have implemented PAWN and the first example discussed in ¹ in MATLAB. The latest code can be downloaded off the MATLAB File Exchange. (It is also available on my Github at https://www.github.com/sriki18/pawn).

The first example discussed in ¹ is the sensitivity of the Ishigami-Homma function. I have included the code for this in ishigami_homma.m and also included the code to reproduce Fig. 4 from ¹. Below is the figure I reproduced (the original is here):

Giving credit

If you found this useful, please cite as :

Srikiran Chandrasekaran (2018). PAWN Global Sensitivity Analysis algorithm (https://www.github.com/sriki18/pawn), GitHub.

Additional information

The original authors of the PAWN paper maintain a website which provides the code for PAWN and additionally a cool sensitivity analysis toolbox for MATLAB called SAFE Toolbox.

References

1. Pianosi, F., Wagener, T., 2015. A simple and efficient method for global sensitivity analysis based on cumulative distribution functions. Environ. Model. Softw. 67, 1–11. https://doi.org/10.1016/j.envsoft.2015.01.004 ↩ ↩² ↩³ ↩⁴