Figure Folders¶

Summary¶

Figure folders contain images written by PyFLASH plotting and pipeline functions. Most ordinary plots save editable-text SVG files through PyFLASH.utils.save_fig(...). A few outputs are PNGs: normality Q-Q diagnostics, pipeline overview montages, and model-sweep summary plots.

Use figure folders when you need publication figures, quick visual inspection, or a way to locate plot-adjacent CSV statistics.

Created By¶

Function or workflow	What it writes
Plot functions such as `plot_mean_bars`, `plot_regressions`, `plot_matrices`, and `plot_images` with `save=True`	SVG figures below `experiment.fig_path` or `batch.fig_path`.
`PyFLASH.utils.save_fig(...)`	The shared save helper used by most plot functions. It creates the destination folder only when saving is enabled.
Pipeline functions such as `correlation`, `adjusted_correlation`, `data_overview`, `group_comparison`, `linear_model`, and `rhythm`	SVG figures inside a named pipeline run folder, plus `! Overview Montage.png` when montage capture is enabled.
Statistical plots routed through `multipleComparisons(..., save_normality=True)`	A `<metric>_normality.png` diagnostic beside the figure/stat CSV for that plot.
`iterative_model_sweep` with `plot=True`	PNG model-sweep summary figures inside the sweep output folder.

Folder Layout¶

For imported batches, the default figure root is:

<batch-output>/Results/Python Figures/

For table-based or custom objects, fig_path can point somewhere else, but the same rules apply.

General plot functions usually create a subfolder from the plot family and, for single-marker plots, the marker name. The helper that builds plot subfolders uses patterns such as:

<fig_path>/
  Bars/
    Period(h).svg
    Period(h).csv
    Period(h)_normality.png
  Iba1/
    Bars/
      Iba1_Count.svg
  Matrices/
    Marker Correlation Matrix.svg

Pipeline figures are grouped by pipeline and run label:

<fig_path>/
  Correlation Pipeline/
    scn_correlations/
      ! Overview Montage.png
      Matrices/
        Pearson Correlation Matrix.svg
        Pearson PValue Matrix.svg
        gate_matrix.csv

Specificity and group labels are encoded into file names rather than deep folders, for example Pearson Correlation Matrix_Diagnosis.AD.svg.

File Contents¶

File type	Contents
`.svg`	Vector Matplotlib figure. PyFLASH sets `svg.fonttype='none'` at the save point so labels, ticks, legends, p-value annotations, and matrix text stay editable as text. Dense scatter/collection artists may be rasterized inside the SVG for file size.
`.png`	Raster figure. Used for normality diagnostics, pipeline montages, and model-sweep summaries.
Plot-side `.csv`	Statistics or data used by a specific plot, when the plot family writes one. For example, bar-chart statistics can sit beside the bar SVG.
`Matrices/*.csv`	Matrix values saved by pipeline matrix figures, such as coefficients, p-values, q-values, and gate masks.

save_fig(...) sanitizes image names with strip_name(...), appends .svg, uses the optional subfolder=..., and returns the full path it attempted to save.

How To Reuse¶

List saved SVG files after a plotting session:

from pathlib import Path

for path in Path(batch.fig_path).rglob("*.svg"):
    print(path)

Find the newest pipeline montage:

from pathlib import Path

montages = sorted(Path(batch.fig_path).rglob("! Overview Montage.png"))
print(montages[-1] if montages else "No montage found")

Read a plot-side CSV, if the plot wrote one:

import pandas as pd

stats = pd.read_csv("analysis-output/Results/Python Figures/Bars/Period(h).csv")
print(stats.head())

Notes¶

save=False on a plot call should avoid writing figures for that call.
Config.SAVE_MODE=False disables writes inside save_fig(...), but the function still returns the path it would use.
Config.SKIP_EXISTING=True makes save_fig(...) skip an existing file unless the caller passes skip_existing=False.
Long Windows paths are handled with an extended-length path prefix. PyFLASH warns when the visible figure path is near the usual Windows path-length limit.
Plot functions and pipelines may create additional subfolders, but the shared rules are: figures live below fig_path, output names are sanitized, and row-filter/group tags go into file names.