raise a
# Load the largest one and make plots
model = comparison_model
with open('exp10-99174-data.pkl', 'rb') as fp:
_, X, mask = pickle.load(fp)
# For the latent space we will just use a corner plot.
component_cmap = mpl_utils.discrete_cmap(16, base_cmap="tab20")
fig = mpl_utils.plot_latent_space(model, X, ellipse_kwds=dict(alpha=0), s=10, edgecolor="none", alpha=1, c=[component_cmap(_) for _ in np.argmax(model.tau_, axis=1)], show_ticks=True,
label_names=[r"$\mathbf{{S}}_{{{0}}}$".format(i + 1) for i in range(model.n_latent_factors)])
for ax in fig.axes:
if ax.is_last_row():
ax.set_ylim(-1, 1)
ax.set_yticks([-1, 0, 1])
fig.tight_layout()
fig.savefig("exp10-size-all-latent-space.pdf", dpi=300)
# Get X_H from mask
import galah_dr2 as galah
elements = [ea.split("_")[0].title() for ea in label_names]
X_H, label_names = galah.get_unflagged_abundances_wrt_h(elements, mask)
ax.set_xlim(0.5, D + 0.5)
ax.set_ylim(-ylim, +ylim)
ax.set_yticks([-ylim, 0, ylim])
fig_factor_loads.tight_layout()
savefig(fig_factor_loads, "factor_loads")
# Plot the latent space.
cmap = mpl_utils.discrete_cmap(data_kwds["n_components"], "Spectral")
label_names = [
f"$\\mathbf{{S}}_{{{i}}}$" for i in range(data_kwds["n_latent_factors"])
]
fig_latent = mpl_utils.plot_latent_space(model,
Y,
cmap=cmap,
label_names=label_names)
for ax in fig_latent.axes:
if ax.get_visible():
if ax.is_last_row():
ax.xaxis.set_major_locator(MaxNLocator(3))
if ax.is_first_col():
ax.yaxis.set_major_locator(MaxNLocator(3))
xlim = np.max(np.abs(ax.get_xlim()))
ylim = np.max(np.abs(ax.get_ylim()))
ax.set_xlim(-xlim, +xlim)
ax.set_ylim(-ylim, +ylim)
fig_latent.tight_layout()
fig_latent.subplots_adjust(hspace=0, wspace=0)
savefig(fig_latent, "latent")