def testFindClosestCentroid(self):
km = kmeans.KmeansClusterer(4)
centroids,mins,maxes = ClusterData.initializeCentroids(4,self.summaryDatas)
self.assertEquals(4,len(centroids))
for centroid in centroids:
self.assertTrue(centroid.totalDist >= mins[clusterdata.TOTAL_DIST])
self.assertTrue(centroid.totalDist <= maxes[clusterdata.TOTAL_DIST])
self.assertTrue(centroid.avgHR >= mins[clusterdata.AVG_HR])
self.assertTrue(centroid.avgHR <= maxes[clusterdata.AVG_HR])
self.assertTrue(centroid.netGained >= mins[clusterdata.NET_GAINED])
self.assertTrue(centroid.netGained <= maxes[clusterdata.NET_GAINED])
self.assertTrue(centroid.netLost >= mins[clusterdata.NET_LOST])
self.assertTrue(centroid.netLost <= maxes[clusterdata.NET_LOST])
self.assertTrue(centroid.timeSeconds >= mins[clusterdata.TIME])
self.assertTrue(centroid.timeSeconds <= maxes[clusterdata.TIME])
centroids,mins,maxes = ClusterData.initializeCentroids(4,self.summaryDatas)
self.assertEquals(4,len(centroids))
# take the 4th centroid and clone it.
data = copy.copy(centroids[3])
centroid = km.findClosestCentroid(centroids, data,mins,maxes)
'''
self.lap = lap
self.totalDist = totalDist
self.avgHR = avgHR
self.netGained = netGained
self.netLost = netLost
self.timeSeconds = timeSeconds
self.goodRecords = goodRecords
self.badRecords = badRecords
'''
self.assertEquals(centroid.lap,centroids[3].lap)
self.assertEquals(centroid.totalDist,centroids[3].totalDist)
self.assertEquals(centroid.avgHR,centroids[3].avgHR)
self.assertEquals(centroid.netGained,centroids[3].netGained)
self.assertEquals(centroid.netLost,centroids[3].netLost)
self.assertEquals(centroid.timeSeconds,centroids[3].timeSeconds)
self.assertEquals(centroid.goodRecords,centroids[3].goodRecords)
self.assertEquals(centroid.badRecords,centroids[3].badRecords)
self.assertEquals(centroid,centroids[3])
pass
