PCA analysis example

In this example, we perform the PCA dimensionality reduction of a spectra dataset

Import the spectrochempy API package

import spectrochempy as scp

Load and inspect the dataset

dataset = scp.read_omnic("irdata/nh4y-activation.spg")[::5]
dataset

_ = dataset.plot()

PCA on the full spectral range

Create a PCA object and fit the dataset so that the explained variance is greater or equal to 99.9%

pca = scp.PCA(n_components=0.999)
_ = pca.fit(dataset)

The number of fitted components is given by the n_components attribute (We obtain 23 components)

pca.n_components

Transform the dataset to a lower dimensionality using all the fitted components

scores = pca.transform()
scores

Display the results graphically

First we can set some preferences for the plot

prefs = scp.preferences
prefs.lines.markersize = 7

# ScreePlot
_ = pca.plot_scree()

Score Plot first we can set some preferences for the plot

prefs.lines.markersize = 10

_ = pca.plot_score()

Score Plot for 3 PC’s in 3D

_ = pca.plot_score(components=(1, 2, 3))

Displays 4 loadings

_ = pca.loadings[:4].plot(legend=True)

Mask the saturated region and refit

Here we do a masking of the saturated region between 882 and 1280 cm^-1

dataset[
    :, 882.0:1280.0
] = scp.MASKED  # remember: use float numbers for slicing (not integer)
_ = dataset.plot()

Apply the PCA model

pca = scp.PCA(n_components=0.999)
_ = pca.fit(dataset)
pca.n_components

As seen above, now only 4 components instead of 23 are necessary to 99.9% of explained variance.

_ = pca.plot_scree()

Displays the loadings

_ = pca.loadings.plot(legend=True)

Let’s plot the scores

scores = pca.transform()
_ = pca.plot_score((1, 2))

Label the score plot with custom labels

Our dataset has already two columns of labels for the spectra but there are a bit too long for display on plots.

scores.y.labels

So we define some short labels for each component, and add them as a third column:

labels = [lab[:6] for lab in dataset.y.labels[:, 1]]
scores.y.labels = labels  # Note this does not replace previous labels,
# but adds a column.

Display the labeled score plot

Labels are now placed automatically using the adjustText library, which provides collision avoidance between labels and markers. Install adjustText with pip install adjustText for best results; without it, labels are still shown with a simple offset from markers.

_ = pca.plot_score(scores=scores, show_labels=True, labels_column=2)

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

# scp.show()