def test_NNMF():
w1 = scp.NDDataset([[1, 2, 3], [4, 5, 6]])
h1 = scp.NDDataset([[1, 2], [3, 4], [5, 6]])
w2 = scp.NDDataset([[1, 1, 1], [4, 4, 4]])
h2 = scp.NDDataset([[1, 1], [3, 3], [5, 5]])
v = scp.dot(w1, h1)
nnmf = scp.NNMF(v,
w2,
h2,
tol=0.0001,
maxtime=60,
maxiter=100,
verbose=True)
scp.info_("------")
scp.info_(nnmf.C)
scp.info_("------")
scp.info_(nnmf.St)
assert_array_almost_equal(nnmf.C.data,
np.array([[1.4, 2.1, 2.9], [4.4, 5.2, 6.0]]),
decimal=1)
assert_array_almost_equal(nnmf.St.data,
[[0.8, 1.9], [2.9, 3.9], [5.1, 5.9]],
decimal=1)
def test_bugs_units_change():
# check for bug on transmittance conversion
X = scp.NDDataset([0.0, 0.3, 1.3, 5.0], units="absorbance")
# A to T
X1 = X.to("transmittance")
assert_array_equal(X1.data, 10 ** -np.array([0.0, 0.3, 1.3, 5.0]) * 100)
assert X1.title == "transmittance"
# T to abs T
X2 = X1.to("absolute_transmittance")
assert_array_equal(X2.data, 10 ** -np.array([0.0, 0.3, 1.3, 5.0]))
assert X2.title == "transmittance"
# A to abs T
X2b = X.to("absolute_transmittance")
assert_array_equal(X2b, X2)
assert X2b.title == "transmittance"
# abs T to T
X3 = X2.to("transmittance")
assert_array_equal(X3, X1)
assert X3.title == "transmittance"
# T to A
X4 = X3.to("absorbance")
assert_array_almost_equal(X4, X)
assert X4.title == "absorbance"
# abs T to A
X5 = X2.to("absorbance")
assert_array_almost_equal(X5, X)
assert X5.title == "absorbance"
def test_nmr_1D_apodization(NMR_dataset_1D):
dataset = NMR_dataset_1D.copy()
dataset /= dataset.real.data.max() # normalize
lb = 0
arr, apod = dataset.em(lb=lb, retapod=True)
# arr and dataset should be equal as no em was applied
assert_equal(dataset.data, arr.data)
lb = 0.0
gb = 0.0
arr, apod = dataset.gm(lb=lb, gb=gb, retapod=True)
# arr and dataset should be equal as no em was applied
assert_equal(dataset.data, arr.data)
lb = 100
arr, apod = dataset.em(lb=lb, retapod=True)
# arr and dataset should not be equal as inplace=False
assert not np.array_equal(dataset.data, arr.data)
assert_array_almost_equal(apod[1], 0.9987, decimal=4)
# inplace=True
dataset.plot(xlim=(0.0, 6000.0))
dataset.em(lb=100.0 * ur.Hz, inplace=True)
dataset.plot(c="r", data_only=True, clear=False)
# successive call
dataset.em(lb=200.0 * ur.Hz, inplace=True)
dataset.plot(c="g", data_only=True, clear=False)
dataset = NMR_dataset_1D.copy()
dataset.plot()
dataset.em(100.0 * ur.Hz, inplace=True)
dataset.plot(c="r", data_only=True, clear=False)
# first test gm
dataset = NMR_dataset_1D.copy()
dataset /= dataset.real.data.max() # normalize
dataset.plot(xlim=(0.0, 6000.0))
dataset, apod = dataset.gm(lb=-100.0 * ur.Hz,
gb=100.0 * ur.Hz,
retapod=True)
dataset.plot(c="b", data_only=True, clear=False)
apod.plot(c="r", data_only=True, clear=False)
show()
def test_compare_scikit_learn():
X = np.array([[-1, -1], [-2, -1], [-3, -2], [1, 1], [2, 1], [3, 2]])
pcas = sklPCA(n_components=2)
pcas.fit(X)
pca = PCA(NDDataset(X))
pca.printev(n_pc=2)
assert_array_almost_equal(pca.sv.data, pcas.singular_values_)
assert_array_almost_equal(pca.ev_ratio.data, pcas.explained_variance_ratio_ * 100.)
dataset = NDDataset.read('irdata/nh4y-activation.spg')
X = dataset.data
pcas = sklPCA(n_components=5)
pcas.fit(X)
dataset = X.copy()
pca = PCA(NDDataset(dataset))
pca.printev(n_pc=5)
assert_array_almost_equal(pca.sv.data[:5], pcas.singular_values_[:5], 4)
assert_array_almost_equal(pca.ev_ratio.data[:5], pcas.explained_variance_ratio_[:5] * 100., 4)
def test_nddataset_coords_indexer():
dx = np.random.random((10, 100, 10))
coord0 = np.linspace(4000, 1000, 10)
coord1 = np.linspace(0, 60, 10) # wrong length
coord2 = np.linspace(20, 30, 10)
with pytest.raises(ValueError): # wrong length
da = scp.NDDataset(
dx,
coordset=[coord0, coord1, coord2],
title="absorbance",
coordtitles=["wavelength", "time-on-stream", "temperature"],
coordunits=["cm^-1", "s", "K"],
)
coord1 = np.linspace(0, 60, 100)
da = scp.NDDataset(
dx,
coordset=[coord0, coord1, coord2],
title="absorbance",
coordtitles=["wavelength", "time-on-stream", "temperature"],
coordunits=["cm^-1", "s", "K"],
)
assert da.shape == (10, 100, 10)
coords = da.coordset
assert len(coords) == 3
assert_array_almost_equal(
da.coordset[2].data, coord0, decimal=2, err_msg="get axis by index " "failed"
)
# we use almost as SpectroChemPy round the coordinate numbers
assert_array_almost_equal(
da.coordset["wavelength"].data,
coord0,
decimal=2,
err_msg="get axis by title failed",
)
assert_array_almost_equal(
da.coordset["time-on-stream"].data,
coord1,
decimal=4,
err_msg="get axis by title failed",
)
assert_array_almost_equal(da.coordset["temperature"].data, coord2, decimal=3)
da.coordset["temperature"] += 273.15 * ur.K
assert_array_almost_equal(
da.coordset["temperature"].data, coord2 + 273.15, decimal=3
)
def test_compare_scikit_learn():
try:
import_optional_dependency("scikit-learn")
except ImportError:
return
from sklearn.decomposition import PCA as sklPCA
X = np.array([[-1, -1], [-2, -1], [-3, -2], [1, 1], [2, 1], [3, 2]])
pcas = sklPCA(n_components=2)
pcas.fit(X)
pca = PCA(NDDataset(X))
pca.printev(n_pc=2)
assert_array_almost_equal(pca._sv.data, pcas.singular_values_)
assert_array_almost_equal(pca.ev_ratio.data,
pcas.explained_variance_ratio_ * 100.0)
dataset = NDDataset.read("irdata/nh4y-activation.spg")
X1 = dataset.copy().data
pcas = sklPCA(n_components=5, svd_solver="full")
pcas.fit(X1)
X2 = NDDataset(dataset.copy())
pca = PCA(X2)
pca.printev(n_pc=5)
assert_array_almost_equal(pca._sv.data[:5], pcas.singular_values_[:5], 4)
assert_array_almost_equal(pca.ev_ratio.data[:5],
pcas.explained_variance_ratio_[:5] * 100.0, 4)
show()