Customizing a Plot#

The default plot is designed to look good with minimal effort. However, every visual aspect can be adjusted using keyword arguments passed to ds.plot(). This page covers the most common customizations.

Line Appearance#

For 1D and 2D datasets, you can adjust how lines appear:

[1]:
import os
from pathlib import Path

import spectrochempy as scp

TEST_FILE = Path(os.environ.get("TEST_FILE", "irdata/nh4y-activation.spg"))

ds = scp.read(TEST_FILE)

Color#

[2]:
_ = ds.plot(color="red")
../../_images/userguide_plotting_customization_4_0.png

Line Width#

[3]:
_ = ds.plot(linewidth=2)
../../_images/userguide_plotting_customization_6_0.png

The short alias lw=... is accepted as well:

[4]:
_ = ds.plot(lw=2)
../../_images/userguide_plotting_customization_8_0.png

Line Style#

[5]:
_ = ds.plot(linestyle="--")
../../_images/userguide_plotting_customization_10_0.png

The corresponding short alias is ls=...:

[6]:
_ = ds.plot(ls="--")
../../_images/userguide_plotting_customization_12_0.png

Marker#

[7]:
_ = ds.plot(marker="o")
../../_images/userguide_plotting_customization_14_0.png

Marker-related aliases such as ms (marker size) and mew (marker edge width) are normalized automatically:

[8]:
_ = ds.plot(marker="o", ms=5, mew=1.5)
../../_images/userguide_plotting_customization_16_0.png

Axis Limits#

Restrict the displayed range:

[9]:
_ = ds.plot(
    xlim=(2000, 1000)
)  # NB: slicing is also possible here, ds[:,2000.:1000.].plot() would give the same result
../../_images/userguide_plotting_customization_18_0.png
[10]:
_ = ds.plot(ylim=(0, 0.5))
../../_images/userguide_plotting_customization_19_0.png

Figure Size#

Control the figure dimensions (width, height in inches):

[11]:
_ = ds.plot(figsize=(10, 4))
../../_images/userguide_plotting_customization_21_0.png

Titles and Labels#

Set the title and labels in the plotting call:

[12]:
_ = ds.plot(
    title="NH4Y Activation Spectrum",
    xlabel="Wavenumber (cm⁻¹)",
    ylabel="Absorbance (a.u.)",
)
../../_images/userguide_plotting_customization_23_0.png

Grid#

Add a grid for easier reading:

[13]:
_ = ds.plot(grid=True)
../../_images/userguide_plotting_customization_25_0.png

Colormap#

As shown in the overview, change colors using the cmap argument:

[14]:
_ = ds.plot(cmap="viridis")
../../_images/userguide_plotting_customization_27_0.png

The aliases c=... for line color and colormap=... for cmap=... are also accepted:

[15]:
_ = ds[0].plot(c="darkred")
../../_images/userguide_plotting_customization_29_0.png
[16]:
_ = ds.plot_image(colormap="plasma")
../../_images/userguide_plotting_customization_30_0.png

Categorical Colors (cmap=None)#

For line plots, passing cmap=None uses categorical colors instead of a continuous colormap:

[17]:
_ = ds.plot_lines(cmap=None)
../../_images/userguide_plotting_customization_32_0.png

Colormap Normalization#

You can customize how colors map to values using Matplotlib’s normalization classes:

[18]:
import matplotlib as mpl

# Centered norm - useful for data with a natural center (e.g., deviations from mean)
norm = mpl.colors.CenteredNorm()
_ = ds.plot_image(cmap="RdBu_r", norm=norm)
../../_images/userguide_plotting_customization_34_0.png
[19]:
# Log norm - useful for data spanning several orders of magnitude
norm = mpl.colors.LogNorm(vmin=0.01, vmax=1.0)
_ = ds.plot_image(cmap="viridis", norm=norm)
../../_images/userguide_plotting_customization_35_0.png

Combining Customizations#

Most arguments can be combined for a tailored plot:

[20]:
_ = ds.plot(
    xlim=(1500, 4000),
    ylim=(0, 0.8),
    linewidth=1.5,
    linestyle="-",
    grid=True,
    figsize=(10, 5),
)
../../_images/userguide_plotting_customization_37_0.png

Composite Plot Customization#

Analysis and comparison plots such as plot_score, plot_compare, plot_merit, plot_baseline, and plot_parity accept keyword arguments for fine-grained control over visual appearance.

Score Plot: Marker and Transparency#

The plot_score function (accessible via PCA.plot_score or standalone scp.plot_score) supports marker, s (marker size), and alpha (transparency). For a PCA model fitted on the dataset used above:

[21]:
pca = scp.PCA(n_components=3)
pca.fit(ds)
_ = pca.plot_score(marker="s", s=30, alpha=0.7)
../../_images/userguide_plotting_customization_39_0.png

These parameters are forwarded to matplotlib.axes.Axes.scatter.

Compare / Merit Plot: Line Style Parameters#

The plot_compare function accepts domain-specific keyword aliases for style control:

Parameter

Effect

Default

exp_linestyle, calc_linestyle, resid_linestyle

Line style per category

"-", "--", "-"

exp_linewidth, calc_linewidth, resid_linewidth

Line width per category

1.0, 1.6, 1.0

exp_c, calc_c, resid_c

Color per category

Semantic defaults

offset

Vertical separation between groups

None

nb_traces

Number of traces to display

"all"

Example, comparing the dataset with its PCA reconstruction:

[22]:
_ = scp.plot_compare(
    ds,
    pca.inverse_transform(),
    exp_linestyle="-",
    calc_linestyle=":",
    exp_linewidth=1.5,
    calc_linewidth=2.0,
    offset=0.1,
)
../../_images/userguide_plotting_customization_41_0.png

Parity Plot: Scatter Parameters#

The plot_parity function (accessible via pls.plot_parity or standalone scp.plot_parity) passes keyword arguments directly to matplotlib.axes.Axes.scatter:

plot_parity(Y, Y_hat, s=20, marker="o", alpha=0.5, c="tab:blue")

Lifecycle Control for Composites#

All composite functions accept the same lifecycle parameters:

  • ax: Existing axes to plot on (creates new figure if None).

  • clear: Whether to clear the axes before plotting (True by default; set to False to overlay on existing axes).

  • show: Whether to display the figure (True by default).

[23]:
import matplotlib.pyplot as plt

_, (ax1, ax2) = plt.subplots(1, 2, figsize=(12, 4))
_ = pca.plot_score(ax=ax1, show=False)
_ = pca.plot_scree(ax=ax2, show=True)
../../_images/userguide_plotting_customization_43_0.png

The Mental Model#

To summarize:

  • ds.plot() gives you a clean default.

  • Keyword arguments customize a single plot.

  • The returned Axes object provides full Matplotlib control.

  • Composite plots (plot_score, plot_compare, etc.) have their own domain-specific parameters but follow the same ax/clear/show contract.

  • Persistent changes across sessions are handled via scp.preferences (covered elsewhere).

These tools cover most day-to-day plotting needs.