libsvm -> kmeans seems to be changing the last col of the dataset from int to enum and changing the data (to predicted cluster ids?)

added additional type checking with inspect and summary to isolate when the dataset changes

Author: Kevin Normoyle

py/h2o_cmd.py

@@ -350,12 +350,15 @@ def infoFromSummary(summaryResult, noPrint=False, numCols=None, numRows=None):
     if 1==0 and numCols and (len(summaries)!=numCols):
         raise Exception("Expected numCols: %s cols in summary. Got %s" % (numCols, len(summaries)))
 
+    coltypeList = []
     for column in summaries:
         colname = column['colname']
+        # is this always None? unused?
         coltype = column['type']
         nacnt = column['nacnt']
         stats = column['stats']
         stattype = stats['type']
+        coltypeList.append(stattype)
         h2o_exec.checkForBadFP(nacnt, 'nacnt for colname: %s stattype: %s' % (colname, stattype))
 
         if stattype == 'Enum':
@@ -419,6 +422,8 @@ def infoFromSummary(summaryResult, noPrint=False, numCols=None, numRows=None):
             print "hbrk:", hbrk
             print "hcnt:", hcnt
 
+    return coltypeList
+
 def dot():
     sys.stdout.write('.')
     sys.stdout.flush()

py/testdir_multi_jvm/test_KMeans_libsvm_fvec.py

@@ -2,7 +2,6 @@
 import random, sys, time, os
 sys.path.extend(['.','..','../..','py'])
 import h2o, h2o_cmd, h2o_browse as h2b, h2o_import as h2i, h2o_kmeans
-
 class Basic(unittest.TestCase):
     def tearDown(self):
         h2o.check_sandbox_for_errors()
@@ -16,10 +15,41 @@ def setUpClass(cls):
     @classmethod
     def tearDownClass(cls):
         # wait while I inspect things
-        # time.sleep(1500)
+        # h2o.sleep(1500)
         h2o.tear_down_cloud()
 
     def test_KMeans_libsvm_fvec(self):
+
+        # hack this into a function so we can call it before and after kmeans
+        # kmeans is changing the last col to enum?? (and changing the data)
+        def do_summary_and_inspect():
+            # SUMMARY******************************************
+            summaryResult = h2o_cmd.runSummary(key=hex_key)
+            coltypeList = h2o_cmd.infoFromSummary(summaryResult)
+
+            # INSPECT******************************************
+            inspect = h2o_cmd.runInspect(None, parseResult['destination_key'], timeoutSecs=360)
+            h2o_cmd.infoFromInspect(inspect, csvFilename)
+
+            numRows = inspect['numRows']
+            numCols = inspect['numCols']
+
+            # Now check both inspect and summary
+            if csvFilename=='covtype.binary.svm':
+                for k in range(55):
+                    naCnt = inspect['cols'][k]['naCnt']
+                    self.assertEqual(0, naCnt, msg='col %s naCnt %d should be %s' % (k, naCnt, 0))
+                    stype = inspect['cols'][k]['type']
+                    print k, stype
+                    self.assertEqual('Int', stype, msg='col %s type %s should be %s' % (k, stype, 'Int'))
+
+                # summary may report type differently than inspect..check it too!
+                # we could check na here too
+                for i,c in enumerate(coltypeList):
+                    print "column index: %s  column type: %s" % (i, c)
+                    # inspect says 'int?"
+                    assert c=='Numeric', "All cols in covtype.binary.svm should be parsed as Numeric! %s %s" % (i,c)
+
         # just do the import folder once
         # make the timeout variable per dataset. it can be 10 secs for covtype 20x (col key creation)
         # so probably 10x that for covtype200
@@ -42,8 +72,12 @@ def test_KMeans_libsvm_fvec(self):
             ("syn_0_100_1000.svm", "cL", 30, 1),
         ]
 
+        csvFilenameList = [
+            ("covtype.binary.svm", "cC", 30, 1),
+        ]
+
         ### csvFilenameList = random.sample(csvFilenameAll,1)
-        # h2b.browseTheCloud()
+        h2b.browseTheCloud()
         lenNodes = len(h2o.nodes)
 
         firstDone = False
@@ -55,16 +89,9 @@ def test_KMeans_libsvm_fvec(self):
             # PARSE******************************************
             # creates csvFilename.hex from file in importFolder dir 
             parseResult = h2i.import_parse(bucket='home-0xdiag-datasets', path=csvPathname, 
-                hex_key=hex_key, timeoutSecs=2000)
-            print "Parse result['destination_key']:", parseResult['destination_key']
+                hex_key=hex_key, timeoutSecs=2000, doSummary=False)
 
-            # INSPECT******************************************
-            start = time.time()
-            inspect = h2o_cmd.runInspect(None, parseResult['destination_key'], timeoutSecs=360)
-            print "Inspect:", parseResult['destination_key'], "took", time.time() - start, "seconds"
-            h2o_cmd.infoFromInspect(inspect, csvFilename)
-            numRows = inspect['numRows']
-            numCols = inspect['numCols']
+            do_summary_and_inspect()
 
             # KMEANS******************************************
             for trial in range(1):
@@ -87,12 +114,16 @@ def test_KMeans_libsvm_fvec(self):
                 elapsed = time.time() - start
                 print "kmeans end on ", csvPathname, 'took', elapsed, 'seconds.', \
                     "%d pct. of timeout" % ((elapsed/timeoutSecs) * 100)
+
+                do_summary_and_inspect()
+
                 # this does an inspect of the model and prints the clusters
                 h2o_kmeans.simpleCheckKMeans(self, kmeans, **kwargs)
 
+                print "hello"
                 (centers, tupleResultList) = h2o_kmeans.bigCheckResults(self, kmeans, csvPathname, parseResult, 'd', **kwargs)
 
-
+                do_summary_and_inspect()
 
 if __name__ == '__main__':
     h2o.unit_main()