1. Save the Main_FitDistribution_GUI.fig and Main_FitDistribution_GUI.m files at the same directory.
2. Run the Main_FitDistribution_GUI.m file.
3. Load data using the "Load Data" button. The file should be a text file with 1 column containing the data. The attached file "DataExample.txt" contains data in a legal format.
4. choose whether you want to fit continuous of discrete distributions, and whether you want to display the PDFs of the CDFs (with the pop-down menus). The lists of optional distributions are displayed on the left side of the GUI.
5. Press "FIT!"
6. You will see the 4 best fits on the graph, and detailed parameters of the best 4 distributions under the graph. The parameters are:
- Distribution name.
- NLogL - Negative of the log likelihood.
- BIC - Bayesian information criterion.
- AIC - Akaike information criterion.
- AICc - AIC with a correction for finite sample sizes.
- Parameters names.
- Parameters values.
This GUI is based on a code written by
Michael Sheppard from MIT Lincoln Laboratory.
Very usefull tool,but i have some problesms.How to estimate the error of the input data and the distribution which is fitted.may someone help.
Thanks a lot.
Very nice and useful tool.
I'm still a matlab beginner, but maybe somebody can help. What is the selecting criterion on which the tool selects the 4 best fits.
Thanks in advance!
A good one. You may want to consider to use Scott Hirsch's submission which allows you to interactively select variables from the workspace.
It gives you more flexibility than using files.
Very useful stuff, thumbs up !!!
Good Job !!!
Very helpful tool
Updated to include an App file for R2012b