lucas992x / gammaprocesses Goto Github PK

This is the code I use for my mastes thesis. If you find it useful for your research, please cite it!

A stochastic process is a gamma process if its increments are independent and gamma distributed with shape parameter v(t) (non-decreasing and right-continuous with v(0) = 0) and rate parameter u > 0. It can be used to approximate the degradation of objects.

gammaprocesses.py can import data from a dataset (two examples are given here) and compute the parameters, or generate random samples from arbitrary parameters and re-solve to compare results. It uses NumPy, SciPy and Matplotlib, that can be easily installed with pip.

The script expects that the dataset contains inspection times and cumulative amounts of deterioration; if it contains increments of deterioration instead, the utility incrtocumul.py can be used to compute cumulative values and save them in a new text file.

• Formulas that can compute the parameters from a dataset are described by J.M. van Noortwijk in "A Survey of the application of Gamma processes in maintenance".
• Laser GaAs degradation dataset is described by W.Q. Meeker, L.A. Escobar in "Statistical Methods for Reliability Data".
• Fatigue-crack dataset is described by L.A. Rodriguez-Picón et al. in "Degradation modeling based on gamma process models with random effects".