forked from h2oai/h2o-2
-
Notifications
You must be signed in to change notification settings - Fork 0
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Added new tests for KMeans using UCI wines data.
- Loading branch information
1 parent
3f1c027
commit 13eb262
Showing
2 changed files
with
261 additions
and
0 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,83 @@ | ||
| import os, json, unittest, time, shutil, sys, random | ||
| sys.path.extend(['.','..','py']) | ||
|
|
||
| import h2o, h2o_cmd, h2o_glm, h2o_hosts, h2o_kmeans | ||
| import h2o_browse as h2b, h2o_import as h2i | ||
|
|
||
| #uses the wines data from http://archive.ics.uci.edu/ml/datasets/Wine | ||
| #PCA performed to collect data into 2 rows. | ||
| #3 groups, small & easy | ||
|
|
||
|
|
||
| class Basic(unittest.TestCase): | ||
| def tearDown(self): | ||
| h2o.check_sandbox_for_errors() | ||
|
|
||
| @classmethod | ||
| def setUpClass(cls): | ||
| global localhost | ||
| localhost = h2o.decide_if_localhost() | ||
| if (localhost): | ||
| h2o.build_cloud(2,java_heap_GB=5) | ||
| else: | ||
| h2o_hosts.build_cloud_with_hosts() | ||
|
|
||
| @classmethod | ||
| def tearDownClass(cls): | ||
| h2o.tear_down_cloud() | ||
|
|
||
| def test_KMeans_winesPCA(self): | ||
| if localhost: | ||
| csvFilenameList = [ | ||
| #with winesPCA2.csv speciy cols = "1,2" | ||
| ('winesPCA.csv', 480, 'cA'), | ||
| ] | ||
| else: | ||
| # None is okay for key2 | ||
| csvFilenameList = [ | ||
| ('winesPCA.csv', 480,'cA'), | ||
| # ('covtype200x.data', 1000,'cE'), | ||
| ] | ||
|
|
||
| importFolderPath = '/home/ec2-user/h2o/smalldata' | ||
| h2i.setupImportFolder(None, importFolderPath) | ||
| for csvFilename, timeoutSecs, key2 in csvFilenameList: | ||
| csvPathname = importFolderPath + "/" + csvFilename | ||
| # creates csvFilename.hex from file in importFolder dir | ||
| start = time.time() | ||
| parseKey = h2i.parseImportFolderFile(None, csvFilename, importFolderPath, | ||
| timeoutSecs=2000, key2=key2, noise=('JStack', None)) | ||
| print "parse end on ", csvPathname, 'took', time.time() - start, 'seconds' | ||
| h2o.check_sandbox_for_errors() | ||
|
|
||
| inspect = h2o_cmd.runInspect(None, parseKey['destination_key']) | ||
| print "\n" + csvPathname, \ | ||
| " num_rows:", "{:,}".format(inspect['num_rows']), \ | ||
| " num_cols:", "{:,}".format(inspect['num_cols']) | ||
|
|
||
| kwargs = { | ||
| #appears not to take 'cols'? | ||
| 'cols': None, | ||
| 'epsilon': 1e-6, | ||
| 'k': 3 | ||
| } | ||
|
|
||
| start = time.time() | ||
| kmeans = h2o_cmd.runKMeansOnly(parseKey=parseKey, \ | ||
| timeoutSecs=timeoutSecs, retryDelaySecs=2, pollTimeoutSecs=60, **kwargs) | ||
| elapsed = time.time() - start | ||
| print "kmeans end on ", csvPathname, 'took', elapsed, 'seconds.', \ | ||
| "%d pct. of timeout" % ((elapsed/timeoutSecs) * 100) | ||
| h2o_kmeans.simpleCheckKMeans(self, kmeans, **kwargs) | ||
| centers = h2o_kmeans.bigCheckResults(self, kmeans, csvPathname, parseKey, 'd', **kwargs) | ||
| print "Expected centers: [-2.276318, -0.965151], with 59 rows." | ||
| print " [0.0388763, 1.63886039], with 71 rows." | ||
| print " [2.740469, -1.237816], with 48 rows." | ||
| model_key = kmeans['destination_key'] | ||
| kmeansScoreResult = h2o.nodes[0].kmeans_score( | ||
| key = parseKey['destination_key'], model_key = model_key) | ||
| score = kmeansScoreResult['score'] | ||
|
|
||
|
|
||
| if __name__ == '__main__': | ||
| h2o.unit_main() |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,178 @@ | ||
| -3.307420974,-1.439402253 | ||
| -2.203249813,0.332455071 | ||
| -2.509660695,-1.028250724 | ||
| -3.74649719,-2.748618391 | ||
| -1.00607049,-0.867384035 | ||
| -3.041673726,-2.116430917 | ||
| -2.442200515,-1.171545343 | ||
| -2.053643789,-1.604437144 | ||
| -2.503811346,-0.915488474 | ||
| -2.745882383,-0.787217029 | ||
| -3.469948375,-1.298669845 | ||
| -1.749816878,-0.61025577 | ||
| -2.107517287,-0.673805614 | ||
| -3.448429214,-1.127449476 | ||
| -4.300652282,-2.090079711 | ||
| -2.298703833,-1.657875063 | ||
| -2.165845683,-2.320758754 | ||
| -1.893629475,-1.626779928 | ||
| -3.532021672,-2.511259707 | ||
| -2.078658558,-1.058153071 | ||
| -3.115613762,-0.784683607 | ||
| -1.083513606,-0.24106354 | ||
| -2.528092634,0.091582278 | ||
| -1.640361079,0.514826666 | ||
| -1.756620656,0.31625681 | ||
| -0.987294063,-0.938021293 | ||
| -1.77028387,-0.684244961 | ||
| -1.231948785,0.089554419 | ||
| -2.182250468,-0.687629899 | ||
| -2.249762671,-0.190923365 | ||
| -2.493187044,-1.237343437 | ||
| -2.669879637,-1.467733346 | ||
| -1.623998013,-0.052556196 | ||
| -1.897338697,-1.628466725 | ||
| -1.406421181,-0.695971074 | ||
| -1.898470874,-0.176213873 | ||
| -1.380966689,-0.656787136 | ||
| -1.119050703,-0.113788776 | ||
| -1.49796891,0.767267636 | ||
| -2.522684901,-1.79793023 | ||
| -2.580815256,-0.777423286 | ||
| -0.666601591,-0.16948285 | ||
| -3.062168979,-1.15266742 | ||
| -0.460908969,-0.329811773 | ||
| -2.095440943,0.070809177 | ||
| -1.132970204,-1.772108487 | ||
| -2.718931185,-1.187983533 | ||
| -2.813403002,-0.644440709 | ||
| -2.004197251,-1.243521635 | ||
| -2.69987528,-1.747039219 | ||
| -3.205874088,-0.166522256 | ||
| -2.85091773,-0.743182376 | ||
| -3.495743283,-1.608197324 | ||
| -2.218533165,-1.869893255 | ||
| -2.140948465,-1.013891472 | ||
| -2.462383396,-1.325269883 | ||
| -2.733806175,-1.432507851 | ||
| -2.167626307,-1.208789989 | ||
| -3.130549248,-1.726708279 | ||
| 0.925969919,3.064840615 | ||
| 1.538141226,1.377557583 | ||
| 1.831084494,0.827649423 | ||
| -0.030520739,1.259233996 | ||
| -2.044494334,1.919617594 | ||
| 0.607965828,1.90269154 | ||
| -0.897695547,0.761762633 | ||
| -2.242182264,1.879291229 | ||
| -0.182868177,2.420318686 | ||
| 0.810518651,0.219893693 | ||
| -1.970063187,1.399335867 | ||
| 1.567793662,0.882493728 | ||
| -1.653018841,0.954021019 | ||
| 0.723331955,1.060653425 | ||
| -2.555019769,-0.259466626 | ||
| -1.827412658,1.284255467 | ||
| 0.865551287,2.437226063 | ||
| -0.368973574,2.147848153 | ||
| 1.453277523,1.379460477 | ||
| -1.259378295,0.768681173 | ||
| -0.375092282,1.024154387 | ||
| -0.75992026,3.365559969 | ||
| -1.031667756,1.446628971 | ||
| 0.493484695,2.374545219 | ||
| 2.531835081,0.087197384 | ||
| -0.832970436,1.469525197 | ||
| -0.785688283,2.020925733 | ||
| 0.804562581,2.227546749 | ||
| 0.556472882,2.366310351 | ||
| 1.111974296,1.797177569 | ||
| 0.554159612,2.650064521 | ||
| 1.345489817,2.112043648 | ||
| 1.560081802,1.847004344 | ||
| 1.927119438,1.555108682 | ||
| -0.744565612,2.306425559 | ||
| -0.954762094,2.217273771 | ||
| -2.536709431,-0.168797864 | ||
| 0.54242248,0.367888776 | ||
| -1.028149464,2.558352543 | ||
| -2.245574925,1.428711158 | ||
| -1.406249155,2.160098391 | ||
| -0.795475851,2.370262577 | ||
| 0.547985924,2.286678198 | ||
| 0.160720367,1.161207694 | ||
| 0.657938975,2.672422602 | ||
| -0.391250737,2.092828092 | ||
| 1.767513135,1.712457831 | ||
| 0.365237067,2.163251026 | ||
| 1.61611371,1.351770206 | ||
| -0.082303615,2.299747276 | ||
| -1.57383547,1.457921668 | ||
| -1.416573265,1.4142173 | ||
| 0.277918778,1.925137507 | ||
| 1.299479291,0.761025552 | ||
| 0.455786148,2.263031866 | ||
| 0.492795729,1.933590622 | ||
| -0.480718361,3.860892733 | ||
| 0.252177515,2.813555672 | ||
| 0.106926012,1.923496091 | ||
| 2.426168672,1.253604772 | ||
| 0.549539355,2.215910732 | ||
| -0.737541413,1.40499335 | ||
| -1.332562734,-0.252624308 | ||
| 1.173775915,0.662099138 | ||
| 0.461034485,0.616548969 | ||
| -0.975721685,1.441504188 | ||
| 0.096537406,2.104062685 | ||
| -0.038378884,1.263198777 | ||
| 1.592665782,1.204745133 | ||
| 0.478215926,1.933386808 | ||
| 1.787790331,1.147052407 | ||
| 1.323368592,-0.169909936 | ||
| 2.377793365,-0.373528925 | ||
| 2.928678651,-0.263119601 | ||
| 2.140772272,-0.367219071 | ||
| 2.363203176,0.458341882 | ||
| 3.055223151,-0.352418705 | ||
| 3.904738983,-0.154147687 | ||
| 3.925390335,-0.657831569 | ||
| 3.085572089,-0.347861484 | ||
| 2.36779237,-0.291159027 | ||
| 2.770996299,-0.285998107 | ||
| 2.280129313,-0.371460001 | ||
| 2.977235064,-0.487841765 | ||
| 2.368513406,-0.480976939 | ||
| 2.203649295,-1.15678934 | ||
| 2.618235277,-0.561576624 | ||
| 4.268597577,-0.647843475 | ||
| 3.572563599,-1.269122707 | ||
| 2.799167604,-1.566115962 | ||
| 2.891502751,-2.035315627 | ||
| 2.314208871,-2.349737747 | ||
| 2.542658413,-2.039529822 | ||
| 1.807442706,-1.523348762 | ||
| 2.752380507,-2.132915648 | ||
| 2.729451047,-0.408733283 | ||
| 3.59472857,-1.79731421 | ||
| 2.881697075,-1.919803079 | ||
| 3.382614128,-1.308186152 | ||
| 1.045233419,-3.505201937 | ||
| 1.605383688,-2.399868419 | ||
| 3.134289512,-0.736084638 | ||
| 2.233855458,-1.172158769 | ||
| 2.839663426,-0.554479845 | ||
| 2.59019044,-0.696002198 | ||
| 2.941003156,-1.550933966 | ||
| 3.520102479,-0.880044297 | ||
| 2.399342284,-2.585064019 | ||
| 2.920845372,-1.270861999 | ||
| 2.175276584,-2.071693314 | ||
| 2.374230368,-2.581385653 | ||
| 3.202583112,0.250542354 | ||
| 3.667572938,-0.845363176 | ||
| 2.458620325,-2.187627269 | ||
| 3.361043046,-2.21005484 | ||
| 2.594636692,-1.752286361 | ||
| 2.670306845,-2.753132874 | ||
| 2.380302543,-2.29088437 | ||
| 3.199732104,-2.761130747 |