Package reflectometry :: Package reduction :: Module wsolve :: Class LinearModel

Class LinearModel

source code


Model evaluator for linear solution to Ax = y.

Computes a confidence interval (range of likely values for the mean at x) or a prediction interval (range of likely values seen when measuring at x). The prediction interval tells you the width of the distribution at x. This should be the same regardless of the number of measurements you have for the value at x. The confidence interval tells you how well you know the mean at x. It should get smaller as you increase the number of measurements. Error bars in the physical sciences usually show a 1-alpha confidence value of erfc(1/sqrt(2)), representing a 1 sigma standandard deviation of uncertainty in the mean.

Confidence intervals for linear system are given by:

x' p +/- sqrt( Finv(1-a,1,df) var(x' p) )

where for confidence intervals:

var(x' p) = sigma^2 (x' inv(A'A) x)

and for prediction intervals:

var(x' p) = sigma^2 (1 + x' inv(A'A) x)

Stored properties:

DoF = len(y)-len(x) = degrees of freedom
rnorm = 2-norm of the residuals y-Ax
x = solution to the equation Ax = y

Computed properties:

cov = covariance matrix [ inv(A'A); O(n^3) ]
var = parameter variance [ diag(cov); O(n^2)]
std = standard deviation of parameters [ sqrt(var); O(n^2) ]
p = test statistic for chisquare goodness of fit [ chi2.sf; O(1) ]

Methods:

ci(A,sigma=1):  return confidence interval evaluated at A
pi(A,alpha=0.05):  return prediction interval evaluated at A
Instance Methods
 
__init__(self, x=None, DoF=None, SVinv=None, rnorm=None)
x.__init__(...) initializes x; see x.__class__.__doc__ for signature
source code
 
__call__(self, A)
Return the prediction for a linear system at points in the rows of A.
source code
 
ci(self, A, sigma=1)
Compute the calculated values and the confidence intervals for the linear model evaluated at A.
source code
 
pi(self, A, p=0.05)
Compute the calculated values and the prediction intervals for the linear model evaluated at A.
source code

Inherited from object: __delattr__, __getattribute__, __hash__, __new__, __reduce__, __reduce_ex__, __repr__, __setattr__, __str__

Properties
  cov
covariance matrix
  var
result variance
  std
result standard deviation
  p
probability of rejection

Inherited from object: __class__

Method Details

__init__(self, x=None, DoF=None, SVinv=None, rnorm=None)
(Constructor)

source code 
x.__init__(...) initializes x; see x.__class__.__doc__ for signature
Overrides: object.__init__

ci(self, A, sigma=1)

source code 

Compute the calculated values and the confidence intervals for the linear model evaluated at A.

sigma=1 corresponds to a 1-sigma confidence interval

Confidence intervals are sometimes expressed as 1-alpha values, where alpha = erfc(sigma/sqrt(2)).

pi(self, A, p=0.05)

source code 

Compute the calculated values and the prediction intervals for the linear model evaluated at A.

p = 1-alpha = 0.05 corresponds to 95% prediction interval


Property Details

cov

covariance matrix
Get Method:
_cov(self)

var

result variance
Get Method:
_var(self)

std

result standard deviation
Get Method:
_std(self)

p

probability of rejection
Get Method:
_p(self)