ca2142ebba
ENH lots of benchmarks fixes
...
* add a title to all benchmark figures/windows
* "bench" is non-standard slang for benchmarking
* remove calls to pyplot.clf because the previous call
to pyplot.figure sets the current figure
* standardize on "Time (s)" as time axis label
* move some benchmark plot legends to the upper left corner of the plot:
most benchmarks expand from (0,0) up and to the right
* simplified plot titles and moved constant info from legend into title
* fixed typos
* made some labels more viewer-friendly
Fixes #2023 .
2013-06-30 17:06:46 +02:00
85ec0fd1ae
P3K death to the print statement
2013-02-11 23:11:57 +01:00
6bdc90417c
COSMIT: pep8
2011-12-23 10:24:36 -08:00
2d734a1ff7
FIX: broken benchmark
2011-12-19 03:53:53 -05:00
36d0adb8c1
pep8 compliant
2011-12-17 00:48:40 +05:30
78dd8b5626
batch trailing spaces cleanup
2011-09-06 23:12:26 +02:00
a57c2ae933
more scikits.learn => sklearn updates
2011-09-03 12:57:53 +02:00
3fd8292c18
Fixes for broken benchmarks due to API changes in samples_generator.py
2011-08-04 15:09:32 +02:00
6134371d54
MISC: s/larslasso_results/lars_lasso_results
2011-07-26 09:52:32 -04:00
5a12763441
API : renaming LARS to Lars
2011-07-25 21:41:48 -04:00
57da10f228
giving more love to benchmarks (pep8, pyflakes, var names, etc...)
2010-12-11 22:07:20 -05:00
90ac3a0965
factorized out correlated regression dataset utility function and updated
...
benchmarks
2010-12-11 23:16:02 +01:00
0992637535
updated lasso benchmark to showcase the region where LassoLARS is faster than Lasso CD
2010-12-08 16:19:04 +01:00
29d1f2bd1b
ENH : allowing Gram matrix precomputing in Lasso / ElasticNet to
...
speed up fitting when n_samples >> n_features. precompute='auto'
is used to do what's best depending on the dimension of the pb
like in LassoLARS
2010-12-07 18:30:26 -05:00
544258d58b
Update LARS benchmarks with latest API changes.
2010-11-26 17:53:10 +01:00
cec8e33e79
Move benchmarks outside the source tree.
...
Some reasons:
- Avoid current fragmentation.
- Avoid nose failures when importing those modules.
2010-11-26 17:53:06 +01:00
Ken Geis