def simulation(self):
sample = bt.sample_distr(self.nums)
print("Simulation results: ")
print(
"Theoretical mean and simulated mean: {},{}\nTheoretical SE and simulated SE: {},{}"
.format(float('%.4f' % round(st.s_mean(self.nums), 4)),
float('%.4f' % round(st.s_mean(sample), 4)),
bt.sample_stand_err(self.nums),
float('%.4f' % round(st.s_std(sample), 4))))
bt.b_plot(sample)
def sample_distr(nums):
sample_props = []
for _ in range(10000):
sample = resample(nums)
sample_props.append(st.s_mean(sample))
sample_props = np.array(sample_props)
return sample_props
def test_mean(self):
self.assertEqual(st.s_mean(self.nums1), 3)
self.assertEqual(st.s_mean(self.nums2), 5)
self.assertEqual(st.s_mean(self.nums3), 3.4)
self.assertEqual(st.s_mean(self.nums4), 5.0)
self.assertEqual(st.s_mean(self.nums5), 5.4)
self.assertIsNone(st.s_mean(self.nums6))
self.assertIsNone(st.s_mean(self.nums7))
def mean(self):
try:
if type(self.nums) != np.ndarray:
raise (NonNumpy)
if self.nums.shape[0] < 50:
raise (LengthError)
return float('%.4f' % round(st.s_mean(self.nums), 4))
except LengthError:
print("The array of size {} too short".format(self.nums.shape[0]))
except NonNumpy:
print("Inapropriate type, pass numpy array")
def p_zscore(x, n, values):
try:
if values <= 0:
raise (nonNegative)
if n <= 0:
raise (nonNegative)
if x == str(x):
raise (nonText)
z = (x - st.s_mean(p_obs(n, values))) / st.s_std(p_obs(n, values))
return z
except nonNegative:
print("Error: values and n should be greater than 0")
except nonText:
print("Error: not supporting string, please type a number")
def knn(tr, te_row, n_neighbors):
try:
if n_neighbors <= 0:
raise (nonNegative)
if te_row == 0:
raise (nonZero)
distances = list()
for tr_row in tr:
d = ds.e_distance(te_row, tr_row)
distances.append((tr_row, d))
distances.sort(key=lambda tup: tup[1])
neighbors = list()
for i in range(n_neighbors):
neighbors.append(distances[i][0])
return st.s_mean(neighbors)
except nonNegative:
print("Error: n_neighbors should be greater than 0")
except nonZero:
print("Error: wrong input")
def mean(self):
return float('%.4f' % round(st.s_mean(self.nums), 4))
def p_zscore(x, n, values):
z = (x - st.s_mean(p_obs(n, values))) / st.s_std(p_obs(n, values))
return z
