def test_bug_243():
import spectrochempy as scp
D = scp.zeros((10, 100))
x = scp.LinearCoord(offset=0.0, increment=1.0, size=100)
y = scp.LinearCoord(offset=0.0, increment=1.0, size=10)
D.set_coordset(x=x, y=y)
D1 = D[:, 0.:10.]
D2 = D[:, 20.:40.]
D12 = scp.concatenate(D1, D2, axis=1)
# D2.x.data[-1] is 40., as expected, but not D12.x.data[-1]:
assert D12.x.data[-1] == D2.x.data[-1]
def test_concatenate(IR_dataset_2D):
dataset = IR_dataset_2D
dim = "x"
# print(dataset)
s = dataset
s1 = dataset[:, -10:]
s2 = dataset[:, :-10]
# specify axis
s = concatenate(s1, s2, dims=dim)
assert s.units == s1.units
assert s.shape[-1] == (s1.shape[-1] + s2.shape[-1])
assert s.x.size == (s1.x.size + s2.x.size)
assert s.x != dataset.x
s = s.sort(dims=dim, descend=True) #
assert_dataset_almost_equal(s.x, Coord(dataset.x, linear=False), decimal=3)
# default concatenation in the last dimension
s = concatenate(s1, s2)
assert s.units == s1.units
assert s.shape[-1] == (s1.shape[-1] + s2.shape[-1])
assert s.x.size == (s1.x.size + s2.x.size)
assert s.x != dataset.x
s = s.sort(descend=True) #
assert_dataset_almost_equal(s.x, Coord(dataset.x, linear=False), decimal=3)
s1 = dataset[:10]
s2 = dataset[20:]
# check with derived units
s1.to(ur.m, force=True)
s2.to(ur.dm, force=True)
s = concatenate(s1, s2, dims=0)
assert s.units == s1.units
assert s.shape[0] == (s1.shape[0] + s2.shape[0])
assert s.y.size == (s1.y.size + s2.y.size)
s = s.sort(dim="y")
# second syntax
s = s1.concatenate(s2, dims=0)
assert s.units == s1.units
assert s.shape[0] == (s1.shape[0] + s2.shape[0])
assert s.y.size == (s1.y.size + s2.y.size)
# third syntax
s = concatenate((s1, s2), dims=0)
assert s.units == s1.units
assert s.shape[0] == (s1.shape[0] + s2.shape[0])
assert s.y.size == (s1.y.size + s2.y.size)
# coordset
coord_2 = Coord(np.log(s.y.data), title="log_time")
s.set_coordset(y=[s.y, coord_2], x=s.x)
s1 = s[:2]
s2 = s[-5:]
s12 = concatenate(s1, s2, axis=0)
assert (s2["y"].labels[1] == s12["y"].labels[1][-5:]).all()
# authors
s0 = s[0]
s1 = s[1]
s0.author = "sdqe65g4rf"
s2 = concatenate(s0, s1)
assert "sdqe65g4rf" in s2.author and s1.author in s2.author
# titles
s0.title = "new_title"
assert concatenate(s0, s1).title == "new_title"
# incompatible dimensions
s0 = scp.NDDataset(np.zeros((10, 100)))
s1 = scp.NDDataset(np.zeros((10, 100)))
with pytest.raises(DimensionsCompatibilityError):
s0.concatenate(s1[0].squeeze())
with pytest.raises(DimensionsCompatibilityError):
s0.concatenate(s1[:, :50], axis=0)
# incompatible units
s0 = scp.NDDataset(np.zeros((10, 100)), units="V")
s1 = scp.NDDataset(np.zeros((10, 100)), units="A")
with pytest.raises(UnitsCompatibilityError):
scp.concatenate(s0, s1)
s1 = scp.NDDataset(np.ones((10, 100)), units="mV")
s01 = scp.concatenate(s0, s1)
assert s01.data[-1, -1] == 0.001
# -------------------
# Stack
# concatenation using stack
s1 = dataset[:10]
s2 = dataset[-10:]
s = stack(s1, s2)
assert s.units == s1.units
assert s.shape == (2, s1.shape[0], s1.shape[1])
assert s.y.size == s1.y.size
assert s.x.size == s1.x.size
with pytest.warns(DeprecationWarning):
concatenate(s1, s2, force_stack=True)
# If one of the dimensions is of size one, then this dimension is NOT removed before stacking
s0 = dataset[0]
s1 = dataset[1]
ss = stack(s0, s1)
assert s0.shape == (1, 5549)
assert ss.shape == (2, s1.shape[0], s1.shape[1])
# # stack squeezed nD dataset
s0 = dataset[0].copy().squeeze()
assert s0.shape == (5549,)
s1 = dataset[1].squeeze()
assert s1.shape == (5549,)
s = stack(s0, s1)
assert s.shape == (2, 5549)
# # stack squeezed nD dataset
s2 = s1[0:100]
with pytest.raises(DimensionsCompatibilityError):
s = stack(s0, s2)
def test_EFA(IR_dataset_2D):
########################################################################################################################
# Generate a test dataset
# ------------------------------------------------------------------
# 1) simulated chromatogram
# *************************
ntimes = 250
ncomponents = 2
t = scp.LinearCoord.arange(ntimes, units="minutes",
title="time") # time coordinates
c = scp.Coord(range(ncomponents),
title="components") # component coordinates
data = np.zeros((ncomponents, ntimes), dtype=np.float64)
data[0] = asymmetricvoigtmodel().f(t,
ampl=4,
width=10,
ratio=0.5,
asym=0.4,
pos=50.0) # compound 1
data[1] = asymmetricvoigtmodel().f(t,
ampl=5,
width=20,
ratio=0.2,
asym=0.9,
pos=120.0) # compound 2
dsc = scp.NDDataset(data=data, coords=[c, t])
dsc.plot()
show()
########################################################################################################################
# 2) absorption spectra
# **********************
spec = np.array([[2.0, 3.0, 4.0, 2.0], [3.0, 4.0, 2.0, 1.0]])
w = scp.Coord(np.arange(1, 5, 1), units="nm", title="wavelength")
dss = scp.NDDataset(data=spec, coords=[c, w])
dss.plot()
########################################################################################################################
# 3) simulated data matrix
# ************************
dataset = scp.dot(dsc.T, dss)
dataset.data = np.random.normal(dataset.data, 0.2)
dataset.title = "intensity"
dataset.plot()
show()
########################################################################################################################
# 4) evolving factor analysis (EFA)
# *********************************
efa = scp.EFA(dataset)
########################################################################################################################
# Plots of the log(EV) for the forward and backward analysis
#
efa.f_ev.T.plot(yscale="log", legend=efa.f_ev.y.labels)
efa.b_ev.T.plot(yscale="log")
########################################################################################################################
# Looking at these EFA curves, it is quite obvious that only two components
# are really significant, and this corresponds to the data that we have in
# input.
# We can consider that the third EFA components is mainly due to the noise,
# and so we can use it to set a cut of values
n_pc = 2
efa.cutoff = np.max(efa.f_ev[:, n_pc].data)
f2 = efa.f_ev
b2 = efa.b_ev
# we concatenate the datasets to plot them in a single figure
both = scp.concatenate(f2, b2)
both.T.plot(yscale="log")
# TODO: add "legend" keyword in NDDataset.plot()
# ########################################################################################################################
# # Get the abstract concentration profile based on the FIFO EFA analysis
# #
# efa.cutoff = None
# c = efa.get_conc(n_pc)
# c.T.plot()
#
# # scp.show() # uncomment to show plot if needed (not necessary in jupyter notebook)
#
# ds = IR_dataset_2D.copy()
#
# # columns masking
# ds[:, 1230.0:920.0] = MASKED # do not forget to use float in slicing
# ds[:, 5900.0:5890.0] = MASKED
#
# # difference spectra
# ds -= ds[-1]
#
# # column masking for bad columns
# ds[10:12] = MASKED
#
# efa = EFA(ds)
#
# n_pc = 4
# c = efa.get_conc(n_pc)
# c.T.plot()
#
show()
efa.b_ev.T.plot(yscale="log") ######################################################################################################################## # Looking at these EFA curves, it is quite obvious that only two components # are really significant, and this corresponds to the data that we have in # input. # We can consider that the third EFA components is mainly due to the noise, # and so we can use it to set a cut of values n_pc = 2 efa.cutoff = np.max(efa.f_ev[:, n_pc].data) f2 = efa.f_ev b2 = efa.b_ev # we concatenate the datasets to plot them in a single figure both = scp.concatenate(f2, b2) both.T.plot(yscale="log") # TODO: add "legend" keyword in NDDataset.plot() ######################################################################################################################## # Get the abstract concentration profile based on the FIFO EFA analysis # efa.cutoff = None c = efa.get_conc(n_pc) c.T.plot() # scp.show() # uncomment to show plot if needed (not necessary in jupyter notebook)