Skip to content

c++ code to generate various distribution data sets to run models agains

Notifications You must be signed in to change notification settings

liquiddrop/distribution_generator

Repository files navigation

distribution_generator

c++ code to generate various distribution data sets to run models agains

The code takes in inputs for various distributions and generates a set of data then outputs it to csv and gnu plot commands. This is using the boost math and random libs to generate the data.

Here is information for useage:

Distribution Types, 'parameters'

Gaussian-G, 'mean' 'sigma'

Exponential-E, 'rate'

ChiSquared-C, 'degrees of freedom'

LogNormal-L, 'mean' 'sigma'

Example Gaussian with 100 point generated with mean of 5 and sigma of 2, output file name example

G 100 5 2 example

Enter 'Q' to quit

A lot of this code is taken or addapted from the book "Practical C++ Financial Programming" by Carlos Oliveira, Apress, 2015. For more information, visit http://coliveira.net

About

c++ code to generate various distribution data sets to run models agains

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages