def rescale_max_check_test(mat):
"""
Checks if a matrix has only ones as maxes after being rescaled twice.
"""
maxes1, rescaled1 = f.rescale(mat)
maxes2, rescaled2 = f.rescale(rescaled1)
return np.all(maxes2 == np.ones(maxes2.shape))
def rescale_data_check_test(mat):
"""
Checks if a matrix matches its original version after being rescaled and re-multiplied
with its maximum values.
Be aware that numerical errors can cause the test to not pass, which is why the
difference is rounded.
"""
maxes, rescaled = f.rescale(mat)
orig = f.reverse_rescale(maxes, rescaled)
diff = np.round(mat - orig, 10)
return np.all(diff == 0.)
def get_value(self, onset):
elapsed = (float(onset) + self.shift) % self.length
if self.interp == 'linear':
val = float(linear(elapsed / self.length, self.arr))
elif self.interp == 'nearest':
val = float(nearest(elapsed / self.length, self.arr))
if self.bounds:
ymin, ymax = self.bounds
s, y, n = rescale(self.arr, ymin, ymax)
val = (val - y) * s + n
inv_qnt = 1.0 / self.quantize
val = floor(val * inv_qnt) / inv_qnt
return val
import functions as f import matplotlib.pyplot as plt import numpy as np import time as t t0 = t.clock() kplr_id = '008191672' kplr_file = 'kplr008191672-2013011073258_llc.fits' jdadj, obsobject, lightdata = f.openfile(kplr_id, kplr_file) time, flux, flux_err = f.fix_data(lightdata) flux, variance = f.rescale(flux, flux_err) time -= np.median(time) depth = 0.00650010001 width = 0.177046694669 period_interval = np.arange(2.00, 8.0, 0.01) offset_intervals = np.arange(0.00, 7.2, 0.01) #Change to numpy arrays to optimize. # z = [[f.sum_chi_squared(flux, f.box(p,o,depth,width,time),variance) for o in offset_intervals] # for p in period_interval] z = [] for p in period_interval: line = [] for o in offset_intervals: if o < p: line.append(f.sum_chi_squared(flux, f.box(p,o,depth,width,time), variance))
import functions as f
import matplotlib.pyplot as plt
import numpy as np
import time as t
t0 = t.clock()
kplr_id = '008191672'
kplr_file = 'kplr008191672-2013011073258_llc.fits'
jdadj, obsobject, lightdata = f.openfile(kplr_id, kplr_file)
time, flux, flux_err = f.fix_data(lightdata)
flux, variance = f.rescale(flux, flux_err)
time -= np.median(time)
depth = 0.00650010001
width = 0.177046694669
period_interval = np.arange(2.00, 8.0, 0.01)
offset_intervals = np.arange(0.00, 7.2, 0.01)
#Change to numpy arrays to optimize.
# z = [[f.sum_chi_squared(flux, f.box(p,o,depth,width,time),variance) for o in offset_intervals]
# for p in period_interval]
z = []
for p in period_interval:
line = []
for o in offset_intervals:
if o < p:
line.append(
(8.2, 87000, 0),
(4.8, 73000, 0),
(2.2, 42000, 1),
(9.1, 98000, 0),
(6.5, 84000, 0),
(6.9, 73000, 0),
(5.1, 72000, 0),
(9.1, 69000, 1),
(9.8, 79000, 1),
]
data = map(list, data) # change tuples to lists
x = [[1] + row[:2] for row in data] # each element is [1, experience, salary]
y = [row[2] for row in data] # each element is paid_account
rescaled_x = fnc.rescale(x)
beta = estimate_beta(rescaled_x, y)
predictions = [predict(x_i, beta) for x_i in rescaled_x]
"""plt.scatter(predictions, y)
plt.xlabel("predicted")
plt.ylabel("actual")
plt.show()"""
def split_data(data, prob):
"""split data into fractions [prob, 1 - prob]"""
results = [], []
for row in data:
results[0 if random.random() < prob else 1].append(row)
return results
def rescale_max_check_test_2(mat):
"""
Similar to before, checks that maximum value of a matrix after rescaling is exactly 1.
"""
maxes, rescaled = f.rescale(mat)
return np.max(np.abs(rescaled)) == 1.
