paper_2025/python/fig_invariance_thresh-lp_appendix.py

import plotstyle_plt
import numpy as np
import matplotlib.pyplot as plt
from thunderhopper.filetools import search_files
from thunderhopper.modeltools import load_data
from plot_functions import ylabel, ylimits, super_xlabel, title_subplot, time_bar
from color_functions import load_colors, shade_colors
from misc_functions import shorten_species
from IPython import embed

# GENERAL SETTINGS:
mode = ['pure', 'noise'][1]
target_species = [
    'Chorthippus_biguttulus',
    'Chorthippus_mollis',
    'Chrysochraon_dispar',
    # 'Euchorthippus_declivus',
    'Gomphocerippus_rufus',
    'Omocestus_rufipes',
    'Pseudochorthippus_parallelus',
]
data_path = '../data/inv/thresh_lp/condensed/'
save_path = f'../figures/fig_invariance_thresh-lp_{mode}_appendix.pdf'

# ANALYSIS SETTINGS:
exclude_zero = True

# SUBSET SETTINGS:
thresh_rel = np.array([0.5, 1, 3])[0]
kern_specs = np.array([
    [1, 0.008],
    [2, 0.004],
    [3, 0.002],
])[np.array([0, 1, 2])]
n_kernels = kern_specs.shape[0]

# GRAPH SETTINGS:
fig_kwargs = dict(
    figsize=(32/2.54, 16/2.54),
    nrows=n_kernels,
    ncols=len(target_species),
    sharex=True,
    sharey=True,
    gridspec_kw=dict(
        wspace=0.4,
        hspace=0.2,
        left=0.07,
        right=0.98,
        bottom=0.1,
        top=0.95,
    )
)

# PLOT SETTINGS:
species_colors = load_colors('../data/species_colors.npz')
kern_shades = [0, 0.75]
kern_colors = shade_colors((0., 0., 0.), np.linspace(*kern_shades, n_kernels))
line_kwargs = dict(
    lw=2,
    alpha=0.5,
    zorder=2,
)
fill_kwargs = dict(
    alpha=0.3,
    zorder=1,
)
mean_kwargs = dict(
    # c=(0.5,) * 3,
    lw=2,
    alpha=1,
    zorder=3,
    ls='--'
)
mean_colors = {
    'Chorthippus_biguttulus': (1,) * 3,
    'Chorthippus_mollis': (0,) * 3,
    'Chrysochraon_dispar': (0,) * 3,
    'Euchorthippus_declivus': (0,) * 3,
    'Gomphocerippus_rufus': (0,) * 3,
    'Omocestus_rufipes': (0,) * 3,
    'Pseudochorthippus_parallelus': (1,) * 3,
}
kern_kwargs = dict(
    lw=2,
)
inset_bounds = [0.05, 0.6, 0.3, 0.25]
kern_bar_time = 0.05
kern_bar_kwargs = dict(
    dur=kern_bar_time,
    y0=0.1,
    y1=0.2,
    color='k',
    lw=0,
    clip_on=False,
    text_pos=(0.5, -1),
    text_str=f'${int(kern_bar_time * 1000)}\\,\\text{{ms}}$',
    text_kwargs=dict(
        fontsize=12,
        ha='center',
        va='top',
    )
)
xlab = 'scale $\\alpha$'
ylabs = [f'$\\mu_{{f_{i}}}$' for i in range(1, n_kernels + 1)]
xlab_kwargs = dict(
    y=0,
    fontsize=16,
    ha='center',
    va='bottom',
)
ylab_kwargs = dict(
    x=0,
    fontsize=20,
    ha='center',
    va='top',
)
title_kwargs = dict(
    x=0.5,
    yref=0.99,
    ha='center',
    va='top',
    fontsize=16,
    fontstyle='italic',
)
letter_kwargs = dict(
    x=0.005,
    y=0.99,
    fontsize=22,
    ha='left',
    va='top',
)

# Prepare graph:
fig, axes = plt.subplots(**fig_kwargs)
axes[0, 0].set_xscale('log')
axes[0, 0].set_ylim(0, 1)
axes[0, 0].yaxis.set_major_locator(plt.MultipleLocator(0.5))
super_xlabel(xlab, fig, axes[-1, 0], axes[-1, -1], **xlab_kwargs)
insets = []
for ax, ylab in zip(axes[:, 0], ylabs):
    ylabel(ax, ylab, **ylab_kwargs, transform=fig.transFigure)
    insets.append(ax.inset_axes(inset_bounds))

# Run through species:
for i, (species, spec_axes) in enumerate(zip(target_species, axes.T)):
    title_subplot(spec_axes[0], shorten_species(species), ref=fig, **title_kwargs)
    
    # Load species data:
    path = search_files(species, incl=[mode, 'unnormed'], dir=data_path)[0]
    data, config = load_data(path, files=['scales', 'mean_feat', 'sd_feat', 'thresh_rel'])
    scales = data['scales']
    means = data['mean_feat']
    sds = data['sd_feat']

    # Reduce to single threshold:
    ind = np.nonzero(data['thresh_rel'] == thresh_rel)[0][0]
    means = means[:, :, ind, :]
    sds = sds[:, :, ind, :]

    if exclude_zero:
        # Exclude zero scale:
        inds = scales > 0
        scales = scales[inds]
        means = means[inds, :, :]
        sds = sds[inds, :, :]

    # Run through kernels:
    for j, (ax, inset) in enumerate(zip(spec_axes, insets)):
        if i == 0:
            # Indicate kernel waveform:
            inset.plot(config['k_times'], config['kernels'][:, j],
                       c=kern_colors[j], **kern_kwargs)
            inset.set_xlim(config['k_times'][[0, -1]])
            ylimits(config['kernels'], inset, pad=0.05)
            inset.set_title(rf'$k_{{{j+1}}}$', fontsize=15)
            if j == 0:
                time_bar(inset, **kern_bar_kwargs)
            inset.axis('off')

        # Plot recording-specific traces:
        for k in range(means.shape[-1]):
            ax.plot(scales, means[:, j, k], c=species_colors[species], **line_kwargs)
            spread = (means[:, j, k] - sds[:, j, k], means[:, j, k] + sds[:, j, k])
            ax.fill_between(scales, *spread, color=species_colors[species], **fill_kwargs)
        # Plot kernel-specific mean trace:
        ax.plot(scales, means[:, j, :].mean(axis=-1), c=mean_colors[species], **mean_kwargs)

# Save graph:
fig.savefig(save_path)
plt.show()