return X_test, y_test, pvals, keys, time
if __name__ == "__main__":
models = {
'SVM': load_model('models/SVM_transit.h5'),
'MLP': load_model('models/MLP_transit.h5'),
'Wavelet MLP': load_model('models/Wavelet MLP_transit.h5'),
'CNN 1D': load_model('models/CNN 1D_transit.h5'),
}
# load OG data
X_test, y_test, pvals, keys, OGtime = load_data('transit_data_test.pkl',
whiten=True,
NULL=False)
colors = {
'MLP': 'red',
'Wavelet MLP': 'green',
'CNN 1D': 'blue',
'SVM': 'orange'
}
# resolution of data relative original
npts = np.arange(X_test.shape[1] * 0.25, X_test.shape[1] * 2,
30).astype(int)
idxs = np.arange(X_test.shape[1])
# filter by
def generate(N):
try:
X_test, y_test, pvals, keys, time = load_data(
'test_dt{}.pkl'.format(N), whiten=True, NULL=False)
except:
print('Generating Data:', N)
settings = {'ws': 360, 'dt': 360. / N}
hw = (0.5 * N * settings['dt']) / 60. / 24. # half width in days
t = np.linspace(1 - hw, 1 + hw, N)
dt = t.max() - t.min()
# default transit parameters
init = {
'rp': 0.1,
'ar': 12.0, # Rp/Rs, a/Rs
'per': 2,
'inc': 90, # Period (days), Inclination
'u1': 0.5,
'u2': 0, # limb darkening (linear, quadratic)
'ecc': 0,
'ome': 0, # Eccentricity, Arg of periastron
'a0': 1,
'a1': 0, # Airmass extinction terms
'a2': 0,
'tm': 1
} # tm = Mid Transit time (Days)
pgrid_test = {
# TEST data
'rp': (np.array([200, 500, 1000, 2500, 5000, 10000]) /
1e6)**0.5, # transit depth (ppm) -> Rp/Rs
'per': np.linspace(*[2, 4, 5]),
'inc': np.array([86, 87, 90]),
'sig_tol': np.linspace(
*[0.25, 3, 12]), # generate noise based on X times the tdepth
# stellar variability systematics
'phi': np.linspace(*[0, np.pi, 4]),
'A': np.array([250, 500, 1000, 2000]) / 1e6,
'w': np.array([6, 12, 24]) / 24., # periods in days
'PA': [
-4 * dt, 4 * dt, 100
], # doubles amp, zeros amp between min time and max time, 1000=no amplitude change
'Pw':
[-12 * dt, 4 * dt,
100], #-12dt=period halfs, 4dt=period doubles, 1000=no change
}
data = dataGenerator(**{
'pgrid': pgrid_test,
'settings': settings,
'init': init
})
data.generate()
pickle.dump(
{
'keys': data.keys,
'results': data.results,
'time': data.t
}, open('pickle_data/test_dt{}.pkl'.format(N), 'wb'))
X_test, y_test, pvals, keys, time = load_data(
'test_dt{}.pkl'.format(N), whiten=True, NULL=False)
return X_test, y_test, pvals, keys, time
from keras.utils import np_utils
import matplotlib.pyplot as plt
import numpy as np
import pickle
if __name__ == "__main__":
models = {
'SVM': load_model('models/SVM_transit.h5'),
'MLP': load_model('models/MLP_transit.h5'),
'Wavelet MLP': load_model('models/Wavelet MLP_transit.h5'),
'CNN 1D': load_model('models/CNN 1D_transit.h5'),
}
# load DATA
X_test, y_test, pvals, keys, time = load_data('transit_data_test.pkl',
whiten=True)
Xw_test = wavy(X_test)
Xc_test = X_test.reshape((X_test.shape[0], X_test.shape[1], 1))
#y_test = np_utils.to_categorical(y_test, num_classes=2)
tests = {
'MLP': X_test,
'Wavelet MLP': Xw_test,
'CNN 1D': Xc_test,
'SVM': X_test
}
colors = {
'MLP': 'red',
'Wavelet MLP': 'green',
'CNN 1D': 'blue',