Changeset 134

Show
Ignore:
Timestamp:
04/03/08 22:14:41 (9 months ago)
Author:
edsuom
Message:

Renaming pypmc to svpmc to reflect specific direction of the subpackage

Files:

Legend:

Unmodified
Added
Removed
Modified
Copied
Moved

  • projects/AsynCluster/trunk/svpmc/sample.py

    r132 r134  
    2222import scipy as s 
    2323from scipy import random, weave 
     24from scipy.stats.distributions as dists 
    2425 
    2526 
    …… 
    8384    I1 = s.greater(V, x[RI,0]).choose(RI, x[RI,1].astype(int)) 
    8485    return I0[I1] 
     86 
     87 
     88class NormalWalk(object): 
     89    """ 
     90    What is a good approximation to the inverse of the cumulative normal 
     91    distribution function? 
     92     
     93    Mathematically we can write this as:  Find X such that Q(X) = p for 
     94    any 0 < p < 1. (Note: this computes an upper tail probability.) 
     95     
     96    Again, it is not possible to write this as a closed form expression, 
     97    so we resort to approximations. Because of the symmetry of the normal 
     98    distribution, we need only consider 0 < p < 0.5. If you have p > 0.5, 
     99    then apply the algorithm below to q = 1-p, and then negate the value 
     100    for X obtained. (This approximation is also from Abramowitz and 
     101    Stegun.) 
     102     
     103    t = sqrt[ ln(1/p^2) ] 
     104     
     105                 c_0 + c_1*t + c_2*t^2 
     106    X = t -  ------------------------------ 
     107             1 + d_1*t + d_2*t^2 + d_3*t^3 
     108     
     109    c_0 = 2.515517 
     110    c_1 = 0.802853 
     111    c_2 = 0.010328 
     112     
     113    d_1 = 1.432788 
     114    d_2 = 0.189269 
     115    d_3 = 0.001308 
     116     
     117    See Abramowitz and Stegun; Press, et al. 
     118    """ 
     119    def cinv(self, p): 
     120        """ 
     121        The inverse of the cumulative distribution function. 
     122 
     123        Returns the random variate value(s) occurring with probability less 
     124        than or equal to the specified probabilit(ies) from 0.0 - 1.0. 
     125        """ 
     126        # TODO: Write C-weave code to do this lightning-fast with the above 
     127        # approximation. 
     128        if not hasattr(self, 'distObj'): 
     129            self.distObj = dists.norm() 
     130        return self.distObj.ppf(p) 
     131 
     132    def kernel(self, x, sdev): 
     133        """ 
     134        For each current random variate value in I{x}, returns the transition 
     135        kernel... 
     136        """ 

  • projects/AsynCluster/trunk/svpmc/test/project.py

    r133 r134  
    2121 
    2222import os.path 
    23 from pypmc.tseries import TimeSeries 
     23from svpmc.tseries import TimeSeries 
    2424 
    2525def ts(fileName): 

  • projects/AsynCluster/trunk/svpmc/test/test_sample.py

    r132 r134  
    1717 
    1818""" 
    19 Unit tests for pypmc.sample 
     19Unit tests for svpmc.sample 
    2020""" 
    2121 

  • projects/AsynCluster/trunk/svpmc/test/test_tseries.py

    r133 r134  
    1717 
    1818""" 
    19 Unit tests for pypmc.tseries 
     19Unit tests for svpmc.tseries 
    2020""" 
    2121