def dimensional(tx, ty, rx, ry, add=None):
print "pca"
for j in range(tx[1].size):
i = j + 1
print "===" + str(i)
compressor = PCA(n_components = i)
t0 = time()
compressor.fit(tx, y=ty)
newtx = compressor.transform(tx)
runtime=time() - t0
V = compressor.components_
print runtime, V.shape, compressor.score(tx)
distances = np.linalg.norm(tx-compressor.inverse_transform(newtx))
print distances
print "pca done"
print "ica"
for j in range(tx[1].size):
i = j + 1
print "===" + str(i)
compressor = ICA(whiten=True)
t0 = time()
compressor.fit(tx, y=ty)
newtx = compressor.transform(tx)
runtime=time() - t0
print newtx.shape, runtime
distances = np.linalg.norm(tx-compressor.inverse_transform(newtx))
print distances
print "ica done"
print "RP"
for j in range(tx[1].size):
i = j + 1
print "===" + str(i)
compressor = RandomProjection(n_components=i)
t0 = time()
compressor.fit(tx, y=ty)
newtx = compressor.transform(tx)
runtime=time() - t0
shape = newtx.shape
print runtime, shape
print "RP done"
print "K-best"
for j in range(tx[1].size):
i = j + 1
print "===" + str(i)
compressor = best(add, k=i)
t0 = time()
compressor.fit(tx, y=ty.ravel())
newtx = compressor.transform(tx)
runtime=time() - t0
shape = newtx.shape
print runtime, shape
print "K-best done"
def test_pipeline_transform():
# Test whether pipeline works with a transformer at the end.
# Also test pipline.transform and pipeline.inverse_transform
iris = load_iris()
X = iris.data
pca = PCA(n_components=2)
pipeline = Pipeline([('pca', pca)])
# test transform and fit_transform:
X_trans = pipeline.fit(X).transform(X)
X_trans2 = pipeline.fit_transform(X)
X_trans3 = pca.fit_transform(X)
assert_array_almost_equal(X_trans, X_trans2)
assert_array_almost_equal(X_trans, X_trans3)
X_back = pipeline.inverse_transform(X_trans)
X_back2 = pca.inverse_transform(X_trans)
assert_array_almost_equal(X_back, X_back2)
def test_pipeline_transform():
# Test whether pipeline works with a transformer at the end.
# Also test pipline.transform and pipeline.inverse_transform
iris = load_iris()
X = iris.data
pca = PCA(n_components=2)
pipeline = Pipeline([('pca', pca)])
# test transform and fit_transform:
X_trans = pipeline.fit(X).transform(X)
X_trans2 = pipeline.fit_transform(X)
X_trans3 = pca.fit_transform(X)
assert_array_almost_equal(X_trans, X_trans2)
assert_array_almost_equal(X_trans, X_trans3)
X_back = pipeline.inverse_transform(X_trans)
X_back2 = pca.inverse_transform(X_trans)
assert_array_almost_equal(X_back, X_back2)
def pca(X, y, components, max_cluster, num_classes, run_nn=False):
X_train, X_test, y_train, y_test = train_test_split(X,
y, test_size=0.3, train_size=0.7, shuffle=True)
pca_compress = PCA(n_components=components, whiten=True)
pca_compress.fit(X_train, y=y_train)
X_train_new = pca_compress.transform(X_train)
X_test_new = pca_compress.transform(X_test)
X_original = pca_compress.inverse_transform(X_test_new)
loss = ((X_test - X_original)**2).mean()
print("Reconstruction Error " + str(loss))
eigenvalues = pca_compress.explained_variance_
print(eigenvalues)
if run_nn:
mlp_classifier(X_train_new, y_train, 0.3, plot=True, X_test=X_test_new, y_test=y_test)
X_new = np.concatenate((X_train_new, X_test_new), axis=0)
y = np.concatenate((y_train, y_test), axis=0)
kmeans(X_new, y,max_cluster, num_classes, run_nn=run_nn, plot_cluster=True,
reduction_algo='PCA')
expectation_max(X_new, y, max_cluster, num_classes, run_nn=run_nn,
plot_cluster=True, reduction_algo='PCA')
r('predictions_dl <- h2o.predict(dlmodel, test3.hex)')
r('head(predictions_dl)')
## new predictions
pred = r('as.matrix(predictions_dl)')
return var(pred -test)
################################################################
figure()
variances_table = []
for i in range(2,11,1):
pca = PCA(n_components=i)
der = derivatives[train_mask_TL]
pca.fit(der)
X = pca.transform(derivatives[test_mask])
pred_pca_temp = (pca.inverse_transform(X))
#
var_fraction_pca_TL = var(pred_pca_temp-derivatives[test_mask])/var(derivatives[test_mask])
#plot([i], [var(pred_pca_temp-derivatives[test_mask])],'D')
var_fraction_DL_TL = DL( derivatives[train_mask_TL], derivatives[test_mask], i)/var(derivatives[test_mask])
#plot([i], [var_DL_TL ],'Dk')
pca = PCA(n_components=i)
der = derivatives[train_mask_no_TL]
pca.fit(der)
X = pca.transform(derivatives[test_mask])
pred_pca_temp = (pca.inverse_transform(X))
var_fraction_pca_no_TL = var(pred_pca_temp-derivatives[test_mask])/var(derivatives[test_mask])
from numpy import loadtxt, genfromtxt, shape, mean, sort, savetxt, size, array, copy
from pylab import figure
from matplotlib.pyplot import plot, savefig, xlabel, ylabel, scatter, axis, xlim, fill_between, legend, text
from sklearn.decomposition.pca import PCA
data_dir= '../data_all_types/'
out_dir='./plots/'
der = loadtxt(data_dir+'derivatives.dat')
flux = loadtxt(data_dir+'fluxes_not_res.dat.gz')
labels = loadtxt(data_dir+'labels.dat')
spectra_data = genfromtxt(data_dir+'spectra_data.dat',dtype=None)
pca = PCA(n_components=4)
pca.fit(der)
X = pca.transform(der)
pred_PCA = (pca.inverse_transform(X))
pca = PCA(n_components=15)
pca.fit(der)
X = pca.transform(der)
pred_PCA_15PC = (pca.inverse_transform(X))
pred_DL = loadtxt('out_DeepLearning/predictions_120,100,90,50,30,20,4,20,30,50,90,100,120_seed1_dl.dat' )
#range_to_plot=[1,2,3,4,5,6,10]
#range_to_plot=range(300)
#range_to_plot=[]
#labels_to_plot=[]
#for i in range(size(labels)):
# if spectra_data['f3'][i]> 0.2 and spectra_data['f3'][i]<.5:
# range_to_plot.append(i)
