def testSimpleAverageCase2Windows(self):
# S2:avg2
print 'Testing for S2:avg2'
self.f.write('S2:avg2\n')
delta, deltaHalfpoint = createBands(np.min(self.y_vals), np.max(self.y_vals), 4)
value1 = mapToBandedArea(np.mean(self.y_vals[0:9]), 0, delta, deltaHalfpoint)
value2 = mapToBandedArea(np.mean(self.y_vals[10:]), 0, delta, deltaHalfpoint)
self.f.write("S2:avgs= 1:" + str(value1) + " 2:" + str(value2) + "\n")
modified_array = []
for x in range(0, 10):
modified_array.append(value1)
for x in range(10, 20):
modified_array.append(value2)
self.calculateAverageDistance(self.y_vals, modified_array)
self.calculateCorrelationCoeff(self.y_vals, modified_array)
self.f.write('-- Reformed distribution array metrics')
self.outputArrayMetrics(modified_array)
self.f.write('\n')
print ''
def testF1AFM(self):
print "Testing for S7:f1Afm"
self.f.write('S7:f1Afm\n')
x_index = arange(1, 21)
min_value = np.min(self.y_vals)
max_value = np.max(self.y_vals)
avg = np.mean(self.y_vals)
delta_a, deltaHalfpoint_a = createBands(min_value, max_value, 4)
band_ya = mapToBandedArea(avg, min_value, delta_a, deltaHalfpoint_a)
corrected_ya = self.correctValue(band_ya, deltaHalfpoint_a, max_value, min_value)
delta_f, deltaHalfpoint_f = createBands(min_value, max_value, 2)
band_1 = mapToBandedArea(self.y_vals[0], min_value, delta_f, deltaHalfpoint_f)
corrected_1 = self.correctValue(band_1, deltaHalfpoint_f, max_value, min_value)
band_2 = mapToBandedArea(self.y_vals[len(self.y_vals) - 1], min_value, delta_f, deltaHalfpoint_f)
corrected_2 = self.correctValue(band_2, deltaHalfpoint_f, max_value, min_value)
x_first = float(x_index[0])
x_last = float(x_index[len(x_index) - 1])
x_mid = (x_last - x_first) / 2 + x_first
x_mid_sub_x1 = x_mid - x_first
ya_sub_y1 = corrected_ya - corrected_1
y2_sub_ya = corrected_2 - corrected_ya
xm_sub_xmid = x_last - x_mid
self.f.write('Band_ya =' + str(band_ya) + "\n")
self.f.write('Band_1 =' + str(band_1) + "\n")
self.f.write('corr_1 =' + str(corrected_1) + "\n")
self.f.write('corr_ya =' + str(corrected_ya) + "\n")
self.f.write('band_2 =' + str(band_2) + "\n")
self.f.write('corr_2 =' + str(corrected_2) + "\n")
self.f.write('xmid =' + str(x_mid) + "\n")
self.f.write('xmid - self.y_vals =' + str(x_mid_sub_x1) + "\n")
self.f.write('ya - y1 =' + str(ya_sub_y1) + "\n")
self.f.write('y2 - ya =' + str(y2_sub_ya) + "\n")
self.f.write('xm-xmid =' + str(xm_sub_xmid) + "\n")
modified_array = []
for x in range(1, 11):
modified_array.append(ya_sub_y1 * (x - x_first) / x_mid_sub_x1 + corrected_1)
for x in range(11, 21):
modified_array.append(y2_sub_ya * (x - x_mid) / xm_sub_xmid + corrected_ya)
self.calculateAverageDistance(self.y_vals, modified_array)
self.calculateCorrelationCoeff(self.y_vals, modified_array)
self.f.write('-- Reformed distribution array metrics')
self.outputArrayMetrics(modified_array)
self.f.write('\n')
print ''
def testA2L(self):
print "Testing for S5:a2L"
self.f.write('S5:a2L\n')
x_index = arange(1, 21)
xa1 = (x_index[len(x_index) - 1] - x_index[0]) * 0.25 + x_index[0]
xa2 = (x_index[len(x_index) - 1] - x_index[0]) * 0.75 + x_index[0]
delta, deltaHalfpoint = createBands(np.min(self.y_vals), np.max(self.y_vals), 4)
ya1 = mapToBandedArea(np.mean(self.y_vals[0:9]), 0, delta, deltaHalfpoint)
ya2 = mapToBandedArea(np.mean(self.y_vals[10:]), 0, delta, deltaHalfpoint)
self.f.write("xa1 = " + str(xa1) + "\n")
self.f.write("xa1 = " + str(xa2) + "\n")
self.f.write("ya1 = " + str(ya1) + "\n")
self.f.write("ya1 = " + str(ya2) + "\n")
self.f.write("ya2-ya1 = " + str(ya2 - ya1) + "\n")
self.f.write("xa1-xa2 = " + str(xa1 - xa2) + "\n")
modified_array = []
for x in range(1, 21):
modified_array.append((ya2 - ya1) * (x - xa1) / (xa2 - xa1) + ya1)
self.calculateAverageDistance(self.y_vals, modified_array)
self.calculateCorrelationCoeff(self.y_vals, modified_array)
self.f.write('-- Reformed distribution array metrics')
self.outputArrayMetrics(modified_array)
self.f.write('\n')
print ''
def testTrendLine(self):
print 'Testing for S4:TL:pq'
self.f.write('S4:TL:pq\n')
xi = arange(1, 21)
min_value = np.min(self.y_vals)
max_value = np.max(self.y_vals)
delta, deltaHalfpoint = createBands(min_value, max_value, 4)
xi = arange(1, 21)
A = array([xi, ones(20)])
# linearly generated sequence
w = linalg.lstsq(A.T, self.y_vals)[0] # obtaining the parameters
fit = w[0] * xi + w[1]
band_p = mapToBandedArea(fit[0], min_value, delta, deltaHalfpoint)
band_q = mapToBandedArea(fit[len(fit) - 1], min_value, delta, deltaHalfpoint)
corrected_p = self.correctValue(band_p, deltaHalfpoint, max_value, min_value)
corrected_q = self.correctValue(band_q, deltaHalfpoint, max_value, min_value)
yq_sub_yp = corrected_q - corrected_p
xq_sub_xp = len(self.y_vals) - 1
self.f.write("delta = " + str(delta) + "\n")
self.f.write("halfpoint = " + str(deltaHalfpoint) + "\n")
self.f.write('TL:p = ' + str(fit[0]) + "\n")
self.f.write('TL:q = ' + str(fit[len(self.y_vals) - 1]) + "\n")
self.f.write("Band_p = " + str(band_p) + "\n")
self.f.write("Band_p = " + str(band_q) + "\n")
self.f.write("Corr_p =" + str(corrected_p) + "\n")
self.f.write("Corr_q =" + str(corrected_q) + "\n")
self.f.write("yq-yp =" + str(yq_sub_yp) + "\n")
self.f.write("xq-xp =" + str(xq_sub_xp) + "\n")
modified_array = []
for x in range(1, 21):
modified_array.append(yq_sub_yp * (x - xi[0]) / xq_sub_xp + corrected_p)
self.calculateAverageDistance(self.y_vals, modified_array)
self.calculateCorrelationCoeff(self.y_vals, modified_array)
self.f.write('-- Reformed distribution array metrics')
self.outputArrayMetrics(modified_array)
self.f.write('\n')
print ''
def testAverageStepChanges(self):
# S3:TL:kb We want to test the average step-changes example
print 'Testing for S3:TL:kb'
self.f.write('S3:TL:kb\n')
min_value = np.min(self.y_vals)
max_value = np.max(self.y_vals)
delta, deltaHalfpoint = createBands(min_value, max_value, 4)
xi = arange(1, 21)
A = array([xi, ones(20)])
# linearly generated sequence
w = linalg.lstsq(A.T, self.y_vals)[0] # obtaining the parameters
k = w[0]
band_b = mapToBandedArea(w[0], min_value, delta, deltaHalfpoint)
corrected_b = self.correctValue(band_b, deltaHalfpoint, max_value, min_value)
if k >= 0.5:
band_k = 0.75
elif k >= 0:
band_k = 0.25
elif k >= -.5:
band_k = -.25
else:
band_k = -.75
corrected_k = band_k
self.f.write("k = " + str(w[0]) + "\n")
self.f.write("b = " + str(w[1]) + "\n")
self.f.write("delta = " + str(delta) + "\n")
self.f.write("halfpoint = " + str(deltaHalfpoint) + "\n")
self.f.write("Band b = " + str(band_b) + "\n")
self.f.write("Corr b = " + str(corrected_b) + "\n")
self.f.write("Band k = " + str(band_k) + "\n")
self.f.write("Corr k = " + str(corrected_k) + "\n")
modified_array = []
for x in range(1, 21):
modified_array.append(corrected_k * x + corrected_b)
self.calculateAverageDistance(self.y_vals, modified_array)
self.calculateCorrelationCoeff(self.y_vals, modified_array)
self.f.write('-- Reformed distribution array metrics')
self.outputArrayMetrics(modified_array)
self.f.write('\n')
def testF1FM(self):
print "Testing for S6:f1fm"
self.f.write('S6:f1fm\n')
x_index = arange(1, 21)
min_value = np.min(self.y_vals)
max_value = np.max(self.y_vals)
delta, deltaHalfpoint = createBands(min_value, max_value, 4)
band_1 = mapToBandedArea((self.y_vals[0]), min_value, delta, deltaHalfpoint)
band_2 = mapToBandedArea((self.y_vals[len(self.y_vals) - 1]), min_value, delta, deltaHalfpoint)
xa1 = x_index[0]
xa2 = x_index[len(x_index) - 1]
ya1 = self.correctValue(band_1, deltaHalfpoint, max_value, min_value)
ya2 = self.correctValue(band_2, deltaHalfpoint, max_value, min_value)
self.f.write("Band_1 =" + str(band_1) + "\n")
self.f.write("Band_2 =" + str(band_2) + "\n")
self.f.write('xa1 =' + str(xa1) + "\n")
self.f.write('xa2 =' + str(xa2) + "\n")
self.f.write('ya1 =' + str(ya1) + "\n")
self.f.write('ya2 =' + str(ya2) + "\n")
self.f.write("ya2-ya1 = " + str(ya2 - ya1) + "\n")
self.f.write("xa2-xa1 = " + str(xa2 - xa1) + "\n")
modified_array = []
for x in range(1, 21):
modified_array.append((ya2 - ya1) * (x - xa1) / (xa2 - xa1) + ya1)
self.calculateAverageDistance(self.y_vals, modified_array)
self.calculateCorrelationCoeff(self.y_vals, modified_array)
self.f.write('-- Reformed distribution array metrics')
self.outputArrayMetrics(modified_array)
self.f.write('\n')
print ''
def testSimpleAverageCase(self):
#S1:avg
print 'Testing for S1:avg'
delta, deltaHalfpoint = createBands(np.min(self.y_vals), np.max(self.y_vals), 16)
value = mapToBandedArea(np.mean(self.y_vals), 0, delta, deltaHalfpoint)
print "S1:avg=", str(value)
modified_array = np.empty(20)
modified_array.fill(value)
self.f.write('S1:avg\n')
self.f.write('Average = ' + str(value) + '\n')
self.calculateAverageDistance(self.y_vals, modified_array)
self.calculateCorrelationCoeff(self.y_vals, modified_array)
self.f.write('-- Reformed distribution array metrics')
self.outputArrayMetrics(modified_array)
self.f.write('\n')
print ''