https://github.com/sukhbinder/uqtoolkit

Simple Uncertainty Quantification Toolkit. Implements polynomial chaos.
https://github.com/sukhbinder/uqtoolkit

Last synced: 2 days ago
JSON representation

Simple Uncertainty Quantification Toolkit. Implements polynomial chaos.

• Host: GitHub
• URL: https://github.com/sukhbinder/uqtoolkit
• Owner: sukhbinder
• Created: 2017-04-09T10:21:00.000Z (almost 8 years ago)
• Default Branch: master
• Last Pushed: 2017-10-31T18:07:26.000Z (about 7 years ago)
• Last Synced: 2024-11-21T23:53:22.641Z (2 months ago)
• Language: Python
• Size: 6.84 KB
• Stars: 1
• Watchers: 2
• Forks: 1
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# UQToolKit

Simple Uncertainty Quantification Toolkit implemented in pure python from a single file.

Implements polynomial chaos.

## Sample usage

```python

import numpy as np

import matplotlib.pyplot as plt

from uqtoolkit import *

# We will implement a simple model, which squares normally distrubuted random numbers

nsamples=5000

f=np.random.standard_normal(nsamples)

Y=f**2

# Defining Polynomial Chaos (PC) exapansion

polyOrder=2 # Polynomial order

param = polyOrder+1

ndim=1  # No of random variables or dimensions

pc_type='HERMITE'

# Create a PC object

pc=uq_pcset(polyOrder,ndim,pc_type)

# Compute the quadtures

quadrature=uq_quadrature(ndim,param,pc_type)

 

# Evaluate quadrature at the PC

K=uq_getNISP(pc,quadrature)

# Evaluate the model at the quadrature nodes

Ygrid=quadrature['nodes']**2

# Project the response on the PC to compute PC coefs

c=np.dot(K,Ygrid)

# Sample 5000 samples from the pc

U=uq_sample(pc,c,nsamples)

# Mean and Std Dev of response

print 'Mean: ',Y.mean(),U.mean()

print 'St dev: ',Y.std(),U.std()

print 'Mean ',c[0]

print 'stddev ',np.sqrt(c.sum()) 

```