This is development documentation. Some features are not available in the stable release. View stable documentation.

EFA example

In this example, we perform the Evolving Factor Analysis

import spectrochempy as scp

Load and preprocess the dataset

dataset = scp.read_omnic("irdata/nh4y-activation.spg")

Change the time origin

dataset.y -= dataset.y[0]

Mask saturated regions and preview

dataset[:, 1230.0:920.0] = scp.MASKED  # do not forget to use float in slicing
dataset[:, 5997.0:5993.0] = scp.MASKED

_ = dataset.plot_stack(title="NH4_Y activation dataset")

Fit the EFA model

efa1 = scp.EFA()
_ = efa1.fit(dataset)

Forward and backward evolution

Forward evolution of the first 5 components

f = efa1.f_ev[:, :5]
_ = f.T.plot(yscale="log", legend=f.k.labels)

Note the use of coordinate k (component axis) in the expression above. To find the actual names of the coordinates, use the dims attribute:

f.dims

['y', 'k']

Backward evolution

b = efa1.b_ev[:, :5]
_ = b.T[:5].plot(yscale="log", legend=b.k.labels)

Select the number of components and extract concentrations

Use 3 components (the 4th component magnitude serves as the noise cutoff):

efa1.n_components = 3
efa1.cutoff = efa1.f_ev[:, 3].max()

C1 = efa1.transform()
_ = C1.T.plot(title="EFA determined concentrations", legend=C1.k.labels)

The same can be done in one step with fit_transform:

efa2 = scp.EFA(n_components=3)
C2 = efa2.fit_transform(dataset)
assert C1 == C2

Extract the resolved components

St = efa2.get_components()
_ = St.plot(title="components", legend=St.k.labels)

Compare with PCA

pca = scp.PCA(n_components=3)
C3 = pca.fit_transform(dataset)

_ = C3.T.plot(title="PCA scores")

LT = pca.loadings
_ = LT.plot(title="PCA components", legend=LT.k.labels)

Uncomment the following line to display all figures when running the script directly with Python.

# scp.show()