def main():
### redirecting stdout
orig_stdout = sys.stdout
f = file('analyszedata.out', 'w')
sys.stdout = f
### Load file
dataFile = 'example.train.csv'
data = util.loadCsv(dataFile)
data = np.asarray(data)[:,:-1]
# print data
### Get and output feature types
analayzeFeatureType(data)
### Get and output value counts for each feature
getValueCountsAll(data)
print "\n%s The number of distinct values and the corresponding feature IDs." % util.RESULT
print distinctValCntFeatureMap
distinctValCntFeatureCntMap = dict()
for keys, values in distinctValCntFeatureMap.items():
distinctValCntFeatureCntMap[keys] = len(values)
print "\n%s The number of distinct values and # of features." % util.RESULT
print distinctValCntFeatureCntMap
# plot
x = [0] * len(distinctValCntFeatureCntMap)
singley = [0] * len(distinctValCntFeatureCntMap)
accumy = [0] * len(distinctValCntFeatureCntMap)
accumValue = 0
for keys, values in distinctValCntFeatureCntMap.items():
x.append(keys)
singley.append(values)
accumValue += values
accumy.append(accumValue)
plt.plot(x, singley)
plt.xlabel('# of distinct values in a feature')
plt.ylabel('# of features')
plt.show()
plt.plot(x, accumy)
plt.xlabel('# of distinct values in a feature')
plt.ylabel('accumalated # of features')
plt.show()
### redirecting stdout
sys.stdout = orig_stdout
f.close()
def countTargetValues(dataFile):
data = util.loadCsv(dataFile)
targets = data[:,-1]
targets = map(int, targets)
print np.count_nonzero(np.asarray(targets))