def test_agoSeq(self):
pdData = pd.DataFrame({'data':[1,2,3,4,5]})
gb = Grey_Bass.Grey_Bass()
testData = np.array([1,2,3,4,5])
res = gb._agoSeq(testData)
res = gb._agoSeq(pdData.data)
ans = np.array([1,3,6,10,15])
print(res)
print(ans)
testing.assert_array_almost_equal(ans,res)
self.assertRaises(ValueError,gb._agoSeq,1)
self.assertRaises(ValueError,gb._agoSeq,'Not Valid')
def test_NLS(self):
gb = Grey_Bass.Grey_Bass()
testData = np.arange(1000,5000,200)
pdData = pd.DataFrame({'data':[1,2,3,4,5]})
res = gb._NLS(testData)
print(res.x)
res = gb._NLS(pdData.data)
self.assertNotAlmostEqual(gb._internalFactor,0)
self.assertNotAlmostEqual(gb._externalFactor,0)
self.assertGreater(gb._marketSize,0)
self.assertRaises(ValueError,gb._NLS,1)
self.assertRaises(ValueError,gb._NLS,[1])
self.assertRaises(ValueError,gb._NLS,'Not Valid')
def test_whitenisation (self):
pdData = pd.DataFrame({'data':[1,2,3,4,5]})
gb = Grey_Bass.Grey_Bass()
trainData = np.arange(100,200,1)
gb._NLS(trainData)
testData = np.arange(201,300,10)
res = gb.predict(testData, True)
gb.predict(pdData.data,True)
for i in res:
self.assertNotEqual(i,0)
score = 0
for i in range(len(res)):
temp = np.square(res[i] - testData[i])
score += temp
score = np.sqrt(score/len(res))
print(res)
print(score)
def test_predict(self):
pdData = pd.DataFrame({'data':[1,2,3,4,5]})
gb = Grey_Bass.Grey_Bass()
trainData = np.arange(100,200,10)
gb._NLS(trainData)
testData = np.arange(201,300,10)
gb.predict(pdData.data)
res = gb.predict(testData, False)
for i in res:
self.assertNotEqual(i,0)
score = 0
for i in range(len(res)):
temp = np.square(res[i] - testData[i])
score += temp
score = np.sqrt(score/len(res))
print(res)
print(score)
self.assertRaises(ValueError,gb.predict,'Not Valid')
import numpy as np
import scipy as sci
import pandas as pd
from scipy.optimize import least_squares
from GreyBass import Grey_Bass
gb = Grey_Bass.Grey_Bass()
testData = np.array([1, 2, 3, 4, 5])
res = gb._NLS(testData)
########################################
import xgboost as xgb
# read in data
dtrain = xgb.DMatrix('demo/data/agaricus.txt.train')
dtest = xgb.DMatrix('demo/data/agaricus.txt.test')
# specify parameters via map
param = {'max_depth': 2, 'eta': 1, 'silent': 1, 'objective': 'binary:logistic'}
num_round = 2
bst = xgb.train(param, dtrain, num_round)
# make prediction
preds = bst.predict(dtest)
def test_rmse(self):
gb = Grey_Bass.Grey_Bass()
resData = np.arange(100,200,1)
realData = np.arange(101,201,1)
res = gb.rmse(resData,realData)
self.assertRaises(ValueError,gb.rmse,[1,1],[1])