Publication plots with Matplotlib rc defaults

The GUI uses pyqtgraph for interactive plotting. For quick figures, use Export Spectra or Export Composite directly in the GUI. For publication figures, it is often better to send a plot to Matplotlib and let Matplotlib handle fonts, line widths, vector export, and final layout.

This page shows how to set useful Matplotlib defaults with a matplotlibrc file. The goal is that every Matplotlib figure starts with readable labels, sensible tick marks, and export settings that work for papers and presentations.

Screenshot placeholder: pyqtgraph plot right-click menu with Export highlighted.

Screenshot placeholder: pyqtgraph export dialog with Matplotlib selected.

Screenshot placeholder: Matplotlib figure window after export, showing the toolbar save button.

Screenshot placeholder: Matplotlib Figure options dialog opened from the toolbar, showing the Axes and Curves tabs.

When To Use This

Use the GUI export directly when you need a fast result image, a Fiji/ImageJ-compatible TIFF, or a quick PNG/PDF of the visible plot.

Use Matplotlib export when the figure should be polished further before publication:

Task	Recommended route	Reason
Save component maps with LUTs and labels	Export Composite	Preserves GUI colors, labels, scale bars, and Fiji metadata.
Save numerical spectra	Save H as CSV	Keeps the actual component spectra for plotting or analysis elsewhere.
Make a fast spectral panel	Export Spectra	Exports the visible spectral plot directly as PNG or PDF.
Prepare a publication spectral panel	pyqtgraph -> Right-click plot -> Export -> Matplotlib	Lets Matplotlib control fonts, ticks, line widths, and vector output.

Find The rc Location

Matplotlib reads style defaults from a file called matplotlibrc. To see which file your current Python environment uses, run:

python -c "import matplotlib as mpl; print(mpl.matplotlib_fname()); print(mpl.get_configdir())"

The first line shows the active rc file. The second line shows the user config directory.

On Windows, a typical user-level location is:

C:\Users\<user>\.matplotlib\matplotlibrc

If the file does not exist yet, create it as a plain text file named exactly matplotlibrc. Do not save it as matplotlibrc.txt.

Use a user-level matplotlibrc for your personal plotting defaults. Avoid editing the matplotlibrc inside site-packages, because that file belongs to the Matplotlib installation and can be overwritten during updates.

Recommended Publication Default

Paste this into your user-level matplotlibrc file:

# Text and fonts
text.usetex: False
font.family: sans-serif
font.sans-serif: Arial, DejaVu Sans, Liberation Sans
mathtext.fontset: dejavusans

# Figure display and export
figure.figsize: 6.5, 4.2
figure.dpi: 120
figure.constrained_layout.use: True
savefig.dpi: 300
savefig.bbox: tight
savefig.pad_inches: 0.04

# Font sizes
font.size: 11
axes.titlesize: 12
axes.labelsize: 11
xtick.labelsize: 10
ytick.labelsize: 10
legend.fontsize: 9

# Axes frame
axes.spines.top: True
axes.spines.right: True
axes.spines.bottom: True
axes.spines.left: True
axes.linewidth: 0.8
axes.edgecolor: black
axes.xmargin: 0.02
axes.ymargin: 0.04
axes.grid: False

# Lines and markers
lines.linewidth: 1.6
lines.markersize: 4

# Mirror ticks on all sides, but label only bottom and left
xtick.bottom: True
xtick.top: True
xtick.labelbottom: True
xtick.labeltop: False

ytick.left: True
ytick.right: True
ytick.labelleft: True
ytick.labelright: False

xtick.direction: in
ytick.direction: in

xtick.major.size: 5
ytick.major.size: 5
xtick.major.width: 0.8
ytick.major.width: 0.8
xtick.major.pad: 4
ytick.major.pad: 4

xtick.minor.visible: True
ytick.minor.visible: True
xtick.minor.size: 2.5
ytick.minor.size: 2.5
xtick.minor.width: 0.6
ytick.minor.width: 0.6

# Legend
legend.frameon: False
legend.handlelength: 1.8

# Color cycle
axes.prop_cycle: cycler('color', ['tab:blue', 'tab:red', 'tab:green', 'tab:purple', 'tab:orange', 'tab:cyan'])

# Vector export with editable text
pdf.fonttype: 42
ps.fonttype: 42
svg.fonttype: none

This style creates a full plot frame, mirrors ticks on the top and right, keeps numeric labels only on the bottom and left, and uses 300 dpi for saved raster images.

Important Settings

Setting	Recommended value	Why it matters
text.usetex	False	Avoids requiring a full LaTeX installation. Matplotlib mathtext still supports labels such as cm$^{-1}$.
figure.dpi	120	Keeps interactive Matplotlib windows readable without becoming oversized.
savefig.dpi	300	Produces publication-quality PNG output.
savefig.bbox	tight	Removes unnecessary whitespace around saved figures.
axes.spines.top/right	True	Draws a full frame around the plot.
xtick.top, ytick.right	True	Mirrors ticks on the top and right sides.
xtick.labeltop, ytick.labelright	False	Avoids duplicate tick labels.
pdf.fonttype, svg.fonttype	42, none	Keeps text editable in common vector workflows.

Matplotlib Export Workflow

1. Generate or inspect the result in the GUI.
2. Right-click the spectral plot.
3. Choose Export.
4. Select Matplotlib in the export dialog.
5. Use the Matplotlib window to check labels, line widths, ticks, and limits.
6. Save the final figure as PDF, SVG, or PNG from the Matplotlib toolbar.

For a paper figure, prefer PDF or SVG when the downstream program supports vector graphics. Use PNG at 300 dpi when the figure must be rasterized.

Edit The Matplotlib Figure Window

After pyqtgraph sends the plot to Matplotlib, the Matplotlib toolbar provides additional interactive controls. The exact icons depend on the Matplotlib backend, but the common buttons are:

Toolbar control	Use it for
Home, back, forward	Return to previous plot views after zooming or panning.
Pan	Move the visible plot range interactively.
Zoom	Draw a rectangle to zoom into a spectral region.
Configure subplots	Adjust margins and spacing if labels are too close to the figure edge.
Figure options	Edit axis labels, axis limits, curve labels, line colors, line widths, markers, and legend entries.
Save	Export the final figure as PNG, PDF, SVG, or another Matplotlib-supported format.

The Figure options button is especially useful for quick publication cleanup. In the dialog:

• use the Axes tab to change the title, x/y labels, limits, scale, and legend settings;
• use the Curves tab to change curve labels, line style, line width, marker style, marker size, and color;
• when curve labels or legend entries change, enable Regenerate legend and then click Apply;
• click Apply to preview changes before saving the final figure.

This is useful when the analysis result is correct but the exported plot still needs readable labels, a cleaner legend, thicker lines, or marker sizes that match the target figure panel.

The legend refresh step is easy to miss. If you rename curves, change which lines should appear in the legend, or edit legend-related settings, tick Regenerate legend before pressing Apply. Without this, Matplotlib may keep the old legend even though the curve settings were changed.

Screenshot placeholder: Matplotlib toolbar with the Figure options button highlighted.

Screenshot placeholder: Matplotlib Curves tab showing curve label, line style, color, marker, and legend-related edits.

Common Problems

Symptom	Likely cause	Fix
Matplotlib ignores the file	File is named matplotlibrc.txt or is in the wrong folder	Re-run the location command and make sure the file is named exactly matplotlibrc.
The plot opens too large	figure.dpi is too high	Use figure.dpi: 120 and keep savefig.dpi: 300.
Top/right ticks are missing	rc file or plotting code disables mirrored ticks	Set xtick.top: True and ytick.right: True; also check for ax.tick_params(top=False, right=False).
The top/right frame is missing	plotting code hides spines	Set all axes.spines.* values to True; check for ax.spines["top"].set_visible(False).
LaTeX errors appear	text.usetex is enabled without a working LaTeX installation	Use text.usetex: False.