def test_read_matlab():
A = NDDataset.read_matlab(MATLABDATA / "als2004dataset.MAT")
assert len(A) == 6
assert A[3].shape == (4, 96)
A = NDDataset.read_matlab(MATLABDATA / "dso.mat")
assert A.name == "Group sust_base line withoutEQU.SPG"
assert A.shape == (20, 426)
def test_read_opus():
# single file
A = NDDataset.read_opus(OPUSDATA / "test.0000")
assert A.shape == (1, 2567)
assert A[0, 2303.8694].data == pytest.approx(2.72740, 0.00001)
# read contents
p = OPUSDATA / "test.0000"
content = p.read_bytes()
F = NDDataset.read_opus({p.name: content})
assert F.name == p.name
assert F.shape == (1, 2567)
assert NDDataset.read_opus(OPUSDATA / "background.0") is None
def test_write():
nd = scp.read_omnic('irdata/nh4y-activation.spg')
# API write methods needs an instance of a NDDataset as the first argument
with pytest.raises(TypeError):
scp.write()
# the simplest way to save a dataset, is to use the function write with a filename as argument
filename = nd.write('essai.scp')
assert filename == cwd / 'essai.scp'
nd2 = NDDataset.load(filename)
testing.assert_dataset_equal(nd2, nd)
filename.unlink()
# if the filename is omitted, the a dialog is opened to select a name (and a protocol)
filename = nd.write()
assert filename is not None
assert filename.stem == nd.name
assert filename.suffix == '.scp'
filename.unlink()
# # a write protocole can be specified
# filename = nd.write(protocole='json')
# assert filename is not None
# assert filename.stem == nd.name
# assert filename.suffix == '.json'
# filename.unlink()
irdatadir = pathclean(prefs.datadir) / "irdata"
for f in ['essai.scp', 'nh4y-activation.scp']:
if (irdatadir / f).is_file():
(irdatadir / f).unlink()
def test_notebook_basecor_bug():
dataset = NDDataset.read_omnic(
os.path.join('irdata', 'nh4y-activation.spg'))
s = dataset[:, 1260.0:5999.0]
s = s - s[-1]
# Important note that we use floating point number
# integer would mean points, not wavenumbers!
basc = BaselineCorrection(s)
ranges = [[1261.86, 1285.89], [1556.30, 1568.26], [1795.00, 1956.75],
[3766.03, 3915.81], [4574.26, 4616.04], [4980.10, 4998.01],
[5437.52, 5994.70]] # predifined ranges
_ = basc.run(*ranges,
method='multivariate',
interpolation='pchip',
npc=5,
figsize=(6, 6),
zoompreview=4)
# The regions used to set the baseline are accessible using the `ranges`
# attibute:
ranges = basc.ranges
print(ranges)
basc.corrected.plot_stack()
def test_issue_375():
# minimal example
n_pc = 3
color1, color2 = "b", "r"
ratio = NDDataset([1, 2, 3])
cum = ratio.cumsum()
ax1 = ratio.plot_bar(color=color1, title="Scree plot")
assert len(ax1.lines) == 0, "no lines"
assert len(ax1.patches) == 3, "bar present"
ax2 = cum.plot_scatter(color=color2, pen=True, markersize=7.0, twinx=ax1)
assert len(ax2.lines) == 1, "1 lines"
assert len(ax2.patches) == 0, "no bar present on the second plot"
# TODO: Don't know yet how to get the marker present.
ax1.set_title("Scree plot")
show()
def make_samples(force_original=False):
_samples = {
'P350': {
'label': r'$\mathrm{M_P}\,(623\,K)$'
},
# 'A350': {'label': r'$\mathrm{M_A}\,(623\,K)$'},
# 'B350': {'label': r'$\mathrm{M_B}\,(623\,K)$'}
}
for key, sample in _samples.items():
# our data are in our test `datadir` directory.
basename = os.path.join(prefs.datadir, f'agirdata/{key}/FTIR/FTIR')
if os.path.exists(basename + '.scp') and not force_original:
# check if the scp file have already been saved
filename = basename + '.scp'
sample['IR'] = NDDataset.read(filename)
else:
# else read the original zip file
filename = basename + '.zip'
sample['IR'] = NDDataset.read_zip(filename,
only=5,
origin='omnic',
merge=True)
# save
sample['IR'].save()
for key, sample in _samples.items():
basename = os.path.join(prefs.datadir, f'agirdata/{key}/TGA/tg')
if os.path.exists(basename + '.scp') and not force_original:
# check if the scp file have already been saved
filename = basename + '.scp'
sample['TGA'] = NDDataset.read(filename)
else:
# else read the original csv file
filename = basename + '.csv'
ss = sample['TGA'] = NDDataset.read_csv(filename, origin='tga')
ss.squeeze(inplace=True)
# lets keep only data from something close to 0.
s = sample['TGA'] = ss[-0.5:35.0]
# save
s.save()
return _samples
def test_generic_read():
# filename + extension specified
ds = scp.read('wodger.spg')
assert ds.name == 'wodger'
# save with no filename (should save wodger.scp)
path = ds.save()
assert isinstance(path, Path)
assert path.stem == ds.name
assert path.parent == ds.directory
# read should be équivalent to load (but read is a more general function,
dataset = NDDataset.load('wodger.scp')
assert dataset.name == 'wodger'
def test_plot2D_as_3D():
data = NDDataset.read_matlab(
os.path.join('matlabdata', 'als2004dataset.MAT'))
X = data[0]
# X.plot_3D()
X.plot_surface()
X.set_coordset(y=Coord(title='elution time'), x=Coord(title='wavenumbers'))
X.title = 'intensity'
X.plot_surface()
X.plot_surface(colorbar=True)
show()
pass
def test_plot2D():
A = NDDataset.read_omnic("irdata/nh4y-activation.spg")
A.y -= A.y[0]
A.y.to("hour", inplace=True)
A.y.title = "Acquisition time"
A.copy().plot_stack()
A.copy().plot_stack(data_transposed=True)
A.copy().plot_image(style=["sans", "paper"], fontsize=9)
# use preferences
prefs = A.preferences
prefs.reset()
prefs.image.cmap = "magma"
prefs.font.size = 10
prefs.font.weight = "bold"
prefs.axes.grid = True
A.plot()
A.plot(style=["sans", "paper", "grayscale"], colorbar=False)
show()
def test_plot2D_as_3D():
data = NDDataset.read_matlab(
os.path.join("matlabdata", "als2004dataset.MAT"))
X = data[0]
X.plot_surface()
X.set_coordset(
y=Coord(title="elution time", units="s"),
x=Coord(title="wavenumbers", units="cm^-1"),
)
X.title = "intensity"
X.plot_surface()
X.plot_surface(colorbar=True)
show()
pass
def test_plot2D():
A = NDDataset.read_omnic('irdata/nh4y-activation.spg')
A.y -= A.y[0]
A.y.to('hour', inplace=True)
A.y.title = u'Aquisition time'
A.copy().plot_stack()
A.copy().plot_stack(data_transposed=True)
A.copy().plot_image(style=['sans', 'paper'], fontsize=9)
# use preferences
prefs = A.preferences
prefs.reset()
prefs.image.cmap = 'magma'
prefs.font.size = 10
prefs.font.weight = 'bold'
prefs.axes.grid = True
A.plot()
A.plot(style=['sans', 'paper', 'grayscale'], colorbar=False)
show()
pass
def test_write():
nd = scp.read_omnic("irdata/nh4y-activation.spg")
# API write methods needs an instance of a NDDataset as the first argument
with pytest.raises(TypeError):
scp.write()
# the simplest way to save a dataset, is to use the function write with a filename as argument
if (cwd / "essai.scp").exists():
(cwd / "essai.scp").unlink()
filename = nd.write("essai.scp") # should not open a DIALOG
assert filename == cwd / "essai.scp"
assert filename.exists()
# try to write it again
filename = nd.write("essai.scp") # should open a DIALOG to confirm
nd2 = NDDataset.load(filename)
testing.assert_dataset_equal(nd2, nd)
filename.unlink()
# if the filename is omitted, a dialog is opened to select a name (and a protocol)
filename = nd.write()
assert filename is not None
assert filename.stem == nd.name
assert filename.suffix == ".scp"
filename.unlink()
# # a write protocole can be specified
# filename = nd.write(protocole='json')
# assert filename is not None
# assert filename.stem == nd.name
# assert filename.suffix == '.json'
# filename.unlink()
irdatadir = pathclean(prefs.datadir) / "irdata"
for f in ["essai.scp", "nh4y-activation.scp"]:
if (irdatadir / f).is_file():
(irdatadir / f).unlink()
def test_plot_1D():
dataset = NDDataset.read_omnic(
os.path.join("irdata", "nh4y-activation.spg"))
# get first 1D spectrum
nd0 = dataset[0, 1550.0:1600.0]
# plot generic 1D
nd0.plot()
nd0.plot_scatter(plottitle=True)
nd0.plot_scatter(marker="^", markevery=10, title="scatter+marker")
prefs = nd0.preferences
prefs.method_1D = "scatter+pen"
nd0.plot(title="xxxx")
prefs.method_1D = "pen"
nd0.plot(marker="o", markevery=10, title="with marker")
# plot 1D column
col = dataset[:, 3500.0] # note the indexing using wavenumber!
_ = col.plot_scatter()
_ = col.plot_scatter(uselabel=True)
def test_read_carroucell(monkeypatch):
# Before testing we need to download the data locally if not yet done:
# because read carrousel is not designed to download itself.
# Use the read_remote for that:
NDDataset.read_remote("irdata/carroucell_samp", replace_existing=False)
nd = NDDataset.read_carroucell("irdata/carroucell_samp", spectra=(1, 2))
for x in nd:
info_(" " + x.name + ": " + str(x.shape))
assert len(nd) == 11
assert nd[3].shape == (2, 11098)
nd = NDDataset.read_carroucell("irdata/carroucell_samp", spectra=(1, 1))
assert isinstance(nd, NDDataset)
monkeypatch.setattr(spectrochempy.core, "open_dialog", dialog_carroucell)
monkeypatch.setenv("KEEP_DIALOGS", "True")
nd = NDDataset.read_carroucell(spectra=(1, 3))
assert nd[3].shape == (3, 11098)
nd = NDDataset.read_carroucell(spectra=(2, 3), discardbg=False)
assert nd[3].shape == (2, 11098)
def test_plotly2D():
A = NDDataset.read_omnic('irdata/nh4y-activation.spg',
directory=prefs.datadir)
A.y -= A.y[0]
A.y.to('hour', inplace=True)
A.y.title = u'Aquisition time'
def test_read_quadera():
# single file
A = NDDataset.read_quadera('msdata/ion_currents.asc')
assert str(A) == 'NDDataset: [float64] A (shape: (y:16975, x:10))'
def fake_dataset(*args, size=3, **kwargs):
if not args:
ds = NDDataset([range(size)])
else:
ds = NDDataset([[range(4)]], )
return ds
def test_1D():
dataset = NDDataset.read_omnic(
os.path.join(prefs.datadir, 'irdata', 'nh4y-activation.spg'))
# get first spectrum
nd0 = dataset[0]
# plot generic
nd0.plot()
# nd0.plot(output=os.path.join(figures_dir, 'IR_dataset_1D'),
# savedpi=150)
#
# # plot generic style
# nd0.plot(style='poster',
# output=os.path.join(figures_dir, 'IR_dataset_1D_poster'),
# savedpi=150)
#
# # check that style reinit to default
# nd0.plot(output='IR_dataset_1D', savedpi=150)
# # try:
# # assert same_images('IR_dataset_1D.png',
# # os.path.join(figures_dir, 'IR_dataset_1D.png'))
# # except AssertionError:
# # os.remove('IR_dataset_1D.png')
# # raise AssertionError('comparison fails')
# # os.remove('IR_dataset_1D.png')
#
# # try other type of plots
# nd0.plot_pen()
# nd0[:, ::100].plot_scatter()
# nd0.plot_lines()
# nd0[:, ::100].plot_bar()
#
# show()
#
# # multiple
# d = dataset[:, ::100]
# datasets = [d[0], d[10], d[20], d[50], d[53]]
# labels = ['sample {}'.format(label) for label in
# ["S1", "S10", "S20", "S50", "S53"]]
#
# # plot multiple
# plot_multiple(method='scatter',
# datasets=datasets, labels=labels, legend='best',
# output=os.path.join(figures_dir,
# 'multiple_IR_dataset_1D_scatter'),
# savedpi=150)
#
# # plot mupltiple with style
# plot_multiple(method='scatter', style='sans',
# datasets=datasets, labels=labels, legend='best',
# output=os.path.join(figures_dir,
# 'multiple_IR_dataset_1D_scatter_sans'),
# savedpi=150)
#
# # check that style reinit to default
# plot_multiple(method='scatter',
# datasets=datasets, labels=labels, legend='best',
# output='multiple_IR_dataset_1D_scatter',
# savedpi=150)
# try:
# assert same_images('multiple_IR_dataset_1D_scatter',
# os.path.join(figures_dir,
# 'multiple_IR_dataset_1D_scatter'))
# except AssertionError:
# os.remove('multiple_IR_dataset_1D_scatter.png')
# raise AssertionError('comparison fails')
# os.remove('multiple_IR_dataset_1D_scatter.png')
# plot 1D column
col = dataset[:, 3500.] # note the indexing using wavenumber!
_ = col.plot_scatter()
show()
def test_read_quadera():
# single file
A = NDDataset.read_quadera(MSDATA / "ion_currents.asc")
assert str(A) == "NDDataset: [float64] A (shape: (y:16975, x:10))"
def test_read_opus():
# single file
A = NDDataset.read_opus(os.path.join('irdata', 'OPUS', 'test.0000'))
assert A.shape == (1, 2567)
assert A[0, 2303.8694].data == pytest.approx(2.72740, 0.00001)
# using a window path
A1 = NDDataset.read_opus('irdata\\OPUS\\test.0000')
assert A1.shape == (1, 2567)
# single file specified with pathlib
datadir = Path(prefs.datadir)
p = datadir / 'irdata' / 'OPUS' / 'test.0000'
A2 = NDDataset.read_opus(p)
assert A2.shape == (1, 2567)
# multiple files not merged
B = NDDataset.read_opus('test.0000', 'test.0001', 'test.0002', directory=os.path.join('irdata', 'OPUS'))
assert isinstance(B, NDDataset)
assert len(B) == 3
# multiple files merged as the merge keyword is set to true
C = scp.read_opus('test.0000', 'test.0001', 'test.0002', directory=os.path.join('irdata', 'OPUS'), merge=True)
assert C.shape == (3, 2567)
# multiple files to merge : they are passed as a list)
D = NDDataset.read_opus(['test.0000', 'test.0001', 'test.0002'], directory=os.path.join('irdata', 'OPUS'))
assert D.shape == (3, 2567)
# multiple files not merged : they are passed as a list but merge is set to false
E = scp.read_opus(['test.0000', 'test.0001', 'test.0002'], directory=os.path.join('irdata', 'OPUS'), merge=False)
assert isinstance(E, list)
assert len(E) == 3
# read contents
p = datadir / 'irdata' / 'OPUS' / 'test.0000'
content = p.read_bytes()
F = NDDataset.read_opus({
p.name: content
})
assert F.name == p.name
assert F.shape == (1, 2567)
# read multiple contents
lst = [datadir / 'irdata' / 'OPUS' / f'test.000{i}' for i in range(3)]
G = NDDataset.read_opus({p.name: p.read_bytes() for p in lst})
assert len(G) == 3
# read multiple contents and merge them
lst = [datadir / 'irdata' / 'OPUS' / f'test.000{i}' for i in range(3)]
H = NDDataset.read_opus({p.name: p.read_bytes() for p in lst}, merge=True)
assert H.shape == (3, 2567)
# read without filename -> open a dialog
K = NDDataset.read_opus()
# read in a directory (assume homogeneous type of data - else we must use the read function instead)
K = NDDataset.read_opus(datadir / 'irdata' / 'OPUS')
assert K.shape == (4, 2567)
# again we can use merge to avoid stacking of all 4 spectra
J = NDDataset.read_opus(datadir / 'irdata' / 'OPUS', merge=False)
assert isinstance(J, list)
assert len(J) == 4
# single opus file using generic read function
# if the protocol is given it is similar to the read_opus function
F = NDDataset.read(os.path.join('irdata', 'OPUS', 'test.0000'), protocol='opus')
assert F.shape == (1, 2567)
# No protocol?
G = NDDataset.read(os.path.join('irdata', 'OPUS', 'test.0000'))
assert G.shape == (1, 2567)
def test_read():
f = Path('irdata/OPUS/test.0000')
A1 = NDDataset.read_opus(f)
assert A1.shape == (1, 2567)
# single file read with protocol specified
A2 = NDDataset.read(f, protocol='opus')
assert A2 == A1
A3 = scp.read('irdata/nh4y-activation.spg', protocol='omnic')
assert str(A3) == 'NDDataset: [float64] a.u. (shape: (y:55, x:5549))'
# single file without protocol
# inferred from filename
A4 = NDDataset.read(f)
assert A4 == A1
A5 = scp.read('irdata/nh4y-activation.spg')
assert str(A5) == 'NDDataset: [float64] a.u. (shape: (y:55, x:5549))'
# native format
f = A5.save_as('nh4y.scp')
A6 = scp.read('irdata/nh4y.scp')
assert str(A6) == 'NDDataset: [float64] a.u. (shape: (y:55, x:5549))'
A7 = scp.read('nh4y', directory='irdata', protocol='scp')
assert str(A7) == 'NDDataset: [float64] a.u. (shape: (y:55, x:5549))'
A8 = scp.read('nh4y', directory='irdata')
assert str(A8) == 'NDDataset: [float64] a.u. (shape: (y:55, x:5549))'
f.unlink()
# multiple compatible 1D files automatically merged
B = NDDataset.read('test.0000',
'test.0001',
'test.0002',
directory=os.path.join('irdata', 'OPUS'))
assert str(B) == 'NDDataset: [float64] a.u. (shape: (y:3, x:2567))'
assert len(B) == 3
# multiple compatible 1D files not merged if the merge keyword is set to False
C = scp.read('test.0000',
'test.0001',
'test.0002',
directory=os.path.join('irdata', 'OPUS'),
merge=False)
assert isinstance(C, list)
# multiple 1D files to merge
D = NDDataset.read(['test.0000', 'test.0001', 'test.0002'],
directory=os.path.join('irdata', 'OPUS'))
assert D.shape == (3, 2567)
# multiple 1D files not merged : they are passed as a list but merge is set to false
E = scp.read(['test.0000', 'test.0001', 'test.0002'],
directory=os.path.join('irdata', 'OPUS'),
merge=False)
assert isinstance(E, list)
assert len(E) == 3
# read contents
datadir = Path(prefs.datadir)
p = datadir / 'irdata' / 'OPUS' / 'test.0000'
content = p.read_bytes()
F = NDDataset.read({p.name: content})
assert F.name == p.name
assert F.shape == (1, 2567)
# read multiple 1D contents and merge them
lst = [datadir / 'irdata' / 'OPUS' / f'test.000{i}' for i in range(3)]
G = NDDataset.read({p.name: p.read_bytes() for p in lst})
assert G.shape == (3, 2567)
assert len(G) == 3
# read multiple 1D contents awithout merging
lst = [datadir / 'irdata' / 'OPUS' / f'test.000{i}' for i in range(3)]
H = NDDataset.read({p.name: p.read_bytes() for p in lst}, merge=False)
isinstance(H, list)
assert len(H) == 3
filename = datadir / 'wodger.spg'
content = filename.read_bytes()
# change the filename to be sure that the file will be read from the passed content
filename = 'try.spg'
# The most direct way to pass the byte content information
nd = NDDataset.read(filename, content=content)
assert str(nd) == 'NDDataset: [float64] a.u. (shape: (y:2, x:5549))'
# It can also be passed using a dictionary structure {filename:content, ....}
nd = NDDataset.read({filename: content})
assert str(nd) == 'NDDataset: [float64] a.u. (shape: (y:2, x:5549))'
# Case where the filename is not provided
nd = NDDataset.read(content)
assert str(nd) == 'NDDataset: [float64] a.u. (shape: (y:2, x:5549))'
# Try with an .spa file
filename = datadir / 'irdata/subdir/7_CZ0-100 Pd_101.SPA'
content = filename.read_bytes()
filename = 'try.spa'
filename2 = datadir / 'irdata/subdir/7_CZ0-100 Pd_102.SPA'
content2 = filename2.read_bytes()
filename = 'try2.spa'
nd = NDDataset.read({filename: content})
assert str(nd) == 'NDDataset: [float64] a.u. (shape: (y:1, x:5549))'
# Try with only a .spa content
nd = NDDataset.read(content)
assert str(nd) == 'NDDataset: [float64] a.u. (shape: (y:1, x:5549))'
# Try with several .spa content (should be stacked into a single nddataset)
nd = NDDataset.read({filename: content, filename2: content2})
assert str(nd) == 'NDDataset: [float64] a.u. (shape: (y:2, x:5549))'
nd = NDDataset.read(content, content2)
assert str(nd) == 'NDDataset: [float64] a.u. (shape: (y:2, x:5549))'
def test_plotly2D():
A = NDDataset.read_omnic("irdata/nh4y-activation.spg",
directory=prefs.datadir)
A.y -= A.y[0]
A.y.to("hour", inplace=True)
A.y.title = "Acquisition time"
# ---- # Let's change this name # %% proj.name = 'myNMRdata' proj # %% [markdown] # Now we will add a dataset to the project. # # First we read the dataset (here some NMR data) and we give it some name (e.g. 'nmr n°1') # %% datadir = pathclean(prefs.datadir) path = datadir / 'nmrdata' / 'bruker' / 'tests' / 'nmr' nd1 = NDDataset.read_topspin(path / 'topspin_1d', expno=1, remove_digital_filter=True, name="NMR_1D") nd2 = NDDataset.read_topspin(path / 'topspin_2d', expno=1, remove_digital_filter=True, name='NMR_2D') # %% [markdown] # To add it to the project, we use the `add_dataset` function for a single dataset: # %% proj.add_datasets(nd1) # %% [markdown] # or `add_datasets` for several datasets. # %% proj.add_datasets(nd1, nd2) # %% [markdown]
# In the first syntax we load the library into a namespace called `scp` # (we recommend this name, but you can choose whatever # you want - except something already in use): # %% import spectrochempy as scp # SYNTAX 1 nd = scp.NDDataset() # %% [markdown] # or in the second syntax, with a wild `*` import. # %% from spectrochempy import * # SYNTAX 2 nd = NDDataset() # %% [markdown] # With the second syntax, as often in python, the access to objects/functions can be greatly simplified. For example, # we can use directly `NDDataset` without a prefix instead of `scp.NDDataset` # which is the first syntax but there is always a # risk of overwriting some variables or functions already present in the namespace. # Therefore, the first syntax is generally highly recommended. # # Alternatively, you can also load only the objects and function required by your application: # # %% from spectrochempy import NDDataset # SYNTAX 3 nd = NDDataset()
# * `mask`: Data can be partially masked at will
# * `units`: Data can have units, allowing units-aware operations
# * `coordset`: Data can have a set of coordinates, one or several by dimensions
#
# Additional metadata can also be added to the instances of this class through the `meta` properties.
# %% [markdown]
# ## 1D-Dataset (unidimensional dataset)
# %% [markdown]
# In the following example, a minimal 1D dataset is created from a simple list, to which we can add some metadata:
# %%
d1D = NDDataset(
[10.0, 20.0, 30.0],
name="Dataset N1",
author="Blake and Mortimer",
description="A dataset from scratch",
)
d1D
# %% [markdown]
# <div class='alert alert-info'>
# <b>Note</b>
#
# In the above code, run in a notebook, the output of d1D is in html for a nice display.
#
# To get the same effect, from a console script, one can use `print_` (with an underscore) and not the usual python
# function `print`. As you can see below, the `print` function only gives a short summary of the information,
# while the `print_` method gives more detailed output
#
# </div>
def test_importer(monkeypatch, fs):
fs.create_file("/var/data/xx1.txt")
assert os.path.exists("/var/data/xx1.txt")
# mock filesystem
fs.create_dir(DATADIR)
# try to read unexistent scp file
f = DATADIR / "fakedir/fakescp.scp"
with pytest.raises(FileNotFoundError):
read(f)
# make fake file
fs.create_file(f)
monkeypatch.setattr(NDDataset, "load", fake_dataset)
nd = read(f)
assert nd == fake_dataset(f)
nd = read(f.stem, directory=DATADIR / "fakedir/", protocol="scp")
assert nd == fake_dataset(f)
nd = read(f.stem, directory=DATADIR / "fakedir/")
assert nd == fake_dataset(f)
# Generic read without parameters and dialog cancel
monkeypatch.setattr(spectrochempy.core, "open_dialog", dialog_cancel)
monkeypatch.setenv(
"KEEP_DIALOGS",
"True") # we ask to display dialogs as we will mock them.
nd = read()
assert nd is None
# read as class method
nd1 = NDDataset.read()
assert nd1 is None
# NDDataset instance as first arguments
nd = NDDataset()
nd2 = nd.read()
assert nd2 is None
nd = read(default_filter="matlab")
assert nd is None
# Check if Filetype is not known
f = DATADIR / "fakedir/not_exist_fake.fk"
with pytest.raises(TypeError):
read_fake(f)
# Make fake type acceptable
FILETYPES.append(("fake", "FAKE files (*.fk)"))
ALIAS.append(("fk", "fake"))
monkeypatch.setattr("spectrochempy.core.readers.importer.FILETYPES",
FILETYPES)
monkeypatch.setattr("spectrochempy.core.readers.importer.ALIAS", ALIAS)
# Check not existing filename
f = DATADIR / "fakedir/not_exist_fake.fk"
with pytest.raises(FileNotFoundError):
read_fake(f)
# Generic read with a wrong protocol
with pytest.raises(spectrochempy.utils.exceptions.ProtocolError):
read(f, protocol="wrongfake")
# Generic read with a wrong file extension
with pytest.raises(TypeError):
g = DATADIR / "fakedir/otherfake.farfelu"
read(g)
# Mock file
f = DATADIR / "fakedir/fake.fk"
fs.create_file(f)
# specific read_(protocol) function
nd = read_fk(f)
assert nd == fake_dataset()
# should also be a Class function
nd = NDDataset.read_fk(f)
assert nd == fake_dataset()
# and a NDDataset instance function
nd = NDDataset().read_fk(f)
assert nd == fake_dataset()
# single file without protocol inferred from filename
nd = read(f)
assert nd == fake_dataset()
# single file read with protocol specified
nd = read(f, protocol="fake")
assert nd == fake_dataset()
# attribute a new name
nd = read(f, name="toto")
assert nd.name == "toto"
# mock some fake file and assume they exists
f1 = DATADIR / "fakedir/fake1.fk"
f2 = DATADIR / "fakedir/fake2.fk"
f3 = DATADIR / "fakedir/fake3.fk"
f4 = DATADIR / "fakedir/fake4.fk"
f5 = DATADIR / "fakedir/otherdir/otherfake.fk"
f6 = DATADIR / "fakedir/emptyfake.fk" # return None when reader
fs.create_file(f1)
fs.create_file(f2)
fs.create_file(f3)
fs.create_file(f4)
fs.create_file(f5)
fs.create_file(f6)
# l = list(pathclean("/Users/christian/test_data/fakedir").iterdir())
# multiple compatible 1D files automatically merged
nd = read(f1, f2, f3)
assert nd.shape == (3, 3)
nd = read([f1, f2, f3], name="fake_merged")
assert nd.shape == (3, 3)
assert nd.name == "fake_merged"
# multiple compatible 1D files not merged if the merge keyword is set to False
nd = read([f1, f2, f3], names=["a", "c", "b"], merge=False)
assert isinstance(nd, list)
assert len(nd) == 3 and nd[0] == fake_dataset()
assert nd[1].name == "c"
# do not merge inhomogeneous dataset
nd = read([f1, f2, f5])
assert isinstance(nd, list)
# too short list of names. Not applied
nd = read([f1, f2, f3], names=["a", "c"], merge=False)
assert nd[0].name.startswith("NDDataset")
monkeypatch.setattr(spectrochempy.core, "open_dialog", dialog_open)
nd = (
read()
) # should open a dialog (but to selects individual filename (here only simulated)
assert nd.shape == (2, 3)
# read in a directory
monkeypatch.setattr(pathlib.Path, "glob", directory_glob)
# directory selection
nd = read(protocol="fake", directory=DATADIR / "fakedir")
assert nd.shape == (4, 3)
nd = read(protocol="fake", directory=DATADIR / "fakedir", merge=False)
assert len(nd) == 4
assert isinstance(nd, list)
nd = read(listdir=True, directory=DATADIR / "fakedir")
assert len(nd) == 4
assert not isinstance(nd, list)
# if a directory is passed as a keyword, the behavior is different:
# a dialog for file selection occurs except if listdir is set to True
nd = read(directory=DATADIR / "fakedir", listdir=False)
assert nd.shape == (2, 3) # -> file selection dialog
nd = read(directory=DATADIR / "fakedir", listdir=True)
assert nd.shape == (4, 3) # -> directory selection dialog
# read_dir()
nd = read_dir(DATADIR / "fakedir")
assert nd.shape == (4, 3)
nd1 = read_dir()
assert nd1 == nd
nd = read_dir(
directory=DATADIR / "fakedir"
) # open a dialog to eventually select directory inside the specified
# one
assert nd.shape == (4, 3)
fs.create_file(DATADIR / "fakedir/subdir/fakesub1.fk")
nd = read_dir(directory=DATADIR / "fakedir", recursive=True)
assert nd.shape == (5, 3)
# no merging
nd = read_dir(directory=DATADIR / "fakedir", recursive=True, merge=False)
assert len(nd) == 5
assert isinstance(nd, list)
# Simulate reading a content
nd = read({"somename.fk": "a fake content"})
assert nd == fake_dataset(content=True)
nd = read_fake({"somename.fk": "a fake content"})
assert nd == fake_dataset(content=True)
# read multiple contents and merge them
nd = read({
"somename.fk": "a fake content",
"anothername.fk": "another fake content",
"stillanothername.fk": "still another fake content",
})
assert nd.shape == (3, 3)
# do not merge
nd = read(
{
"somename.fk": "a fake content",
"anothername.fk": "another fake content"
},
merge=False,
)
assert isinstance(nd, list)
assert len(nd) == 2
# To load the API, you must import it using one of the following syntax. # # In the first syntax we load the library into a namespace called `scp` (you can choose whatever you want - except # something already in use): # %% import spectrochempy as scp # SYNTAX 1 nd = scp.NDDataset() # %% [markdown] # or in the second syntax, with a wild `*` import. # %% from spectrochempy import * nd = NDDataset() # %% [markdown] # With the second syntax, as often in python, the access to objects/functions can be greatly simplified. For example, # we can use "NDDataset" without a prefix instead of `scp.NDDataset` which is the first syntax) but there is always a # risk of overwriting some variables or functions already present in the namespace. # Therefore, the first syntax is generally highly recommended. # # Alternatively, you can also load only the onbjects and function required by your application: # # %% from spectrochempy import NDDataset nd = NDDataset()
# handling spectroscopic information, one of the major objectives of the SpectroChemPy package: # # * `mask`: Data can be partially masked at will # * `units`: Data can have units, allowing units-aware operations # * `coordset`: Data can have a set of coordinates, one or sevral by dimensions # # Additional metadata can also be added to the instances of this class through the `meta` properties. # %% [markdown] # ## 1D-Dataset (unidimensional dataset) # %% [markdown] # In the following example, a minimal 1D dataset is created from a simple list, to which we can add some metadata: # %% d1D = NDDataset([10., 20., 30.], name="Dataset N1", author='Blake and Mortimer', description='A dataset from scratch') d1D # %% [markdown] # <div class='alert alert-info'> # <b>Note</b> # # In the above code, run in a notebook, the output of d1D is in html for a nice display. # # To get the same effect, from a console script, one can use `print_` (with an underscore) and not the usual python # function `print`. As you can see below, the `print` function only gives a short summary of the information, # while the `print_` method gives more detailed output # # </div> # %%
