baohan
/
scikit-learn
mirror of https://github.com/scikit-learn/scikit-learn.git
0
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

More vectorization of the johnson_lindenstraus_bound function

This commit is contained in:
Olivier Grisel 2012-01-04 11:36:33 +01:00 committed by Andreas Mueller
parent 4ca8e34173
commit b0c1b6c5e5
1 changed files with 5 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
sklearn/random_projection.py
View File

@ -66,11 +66,11 @@ def johnson_lindenstrauss_bound(n_samples, eps=0.1):
>>> johnson_lindenstrauss_bound(1e6, eps=0.5)
663
>>> johnson_lindenstrauss_bound(1e6, eps=0.1)
11841
>>> johnson_lindenstrauss_bound(1e6, eps=[0.5, 0.1, 0.01])
array([ 663, 11841, 1112658])
>>> johnson_lindenstrauss_bound(1e6, eps=0.01)
1112658
>>> johnson_lindenstrauss_bound([1e4, 1e5, 1e6], eps=0.1)
array([ 7894, 9868, 11841])
References
----------
@ -81,6 +81,7 @@ def johnson_lindenstrauss_bound(n_samples, eps=0.1):
http://citeseer.ist.psu.edu/viewdoc/summary?doi=10.1.1.45.3654
"""
eps = np.asarray(eps)
denominator = (eps ** 2 / 2) - (eps ** 3 / 3)
return (4 * np.log(n_samples) / denominator).astype(np.int)