Plotting¶

Publication-quality plotting utilities for SED fits, SFH recovery, corner plot comparisons, and convergence diagnostics tables. Designed for ApJ/MNRAS figures.

Style Setup¶

tengri.analysis.plotting.setup_style(style='tengri')[source]¶

Configure matplotlib for publication-quality astronomy figures.

Defaults to a single cohesive style across every tengri notebook and example, built on top of the scienceplots science preset — the standard Nature/PRL/ApJ-compatible look (serif fonts, thin inward ticks on all four sides, frameless legends, no grid).

Parameters:

style (str) –

Which composite style to apply. One of:

"tengri" (default) — scienceplots.science + no-latex + tengri-specific tick and font-size overrides. Safe everywhere (no system LaTeX required).
"tengri-nature" — Nature-journal variant (single-column width, sans-serif Arial). Use for slide decks.
"tengri-minimal" — same tengri overrides but skip scienceplots so the style works when scienceplots is unavailable. Keeps the BAGPIPES-inspired look as a fallback.

Return type:

None

Examples

>>> from tengri.analysis.plotting import setup_style
>>> setup_style()  # default: science + tengri overrides
>>> setup_style("tengri-minimal")  # fallback if scienceplots absent

SFH Plots¶

tengri.analysis.plotting.plot_sfh(model, posterior, true_params=None, ax=None, color=None, label='Posterior', method='RT', show_draws=True, n_draws=30, ci_levels=(16, 84), xlim=(0, 13.5), show_mean_sfh=True)[source]¶

Plot SFH posterior with uncertainty band — BAGPIPES/Prospector style.

Features: - Filled 68% credible interval (not sample spaghetti) - Optional faint sample draws underneath - Truth as solid black line - Mean SFH backbone as dashed line - Clean axis labels with proper units

Parameters:

model (SEDModel) – Fitted model instance.
posterior (Posterior) – Posterior from Fitter.run().
true_params (dict, optional) – Ground truth parameters for mock recovery plots.
ax (matplotlib Axes, optional) – Axes to plot on. Creates new figure if None.
color (str, optional) – Line color. Defaults to sampler-specific color from style sheet.
label (str) – Legend label. Default "Posterior".
method (str) – Sampler name for auto-coloring ("RT", "geoVI", "NUTS", "MAP").
show_draws (bool) – If True, draw faint individual SFH samples. Default True.
n_draws (int) – Number of sample draws to plot. Default 30.
ci_levels (tuple of float) – Percentile levels for the filled credible interval. Default (16, 84).
xlim (tuple of float) – x-axis limits in lookback time. [Gyr]
show_mean_sfh (bool) – If True, overplot the smooth mean SFH backbone as a dashed line. Default True.

Returns:

Return type:

matplotlib Axes

Examples

from tengri import plot_sfh

ax = plot_sfh(model, posterior, true_params=true_params, method="NUTS")
ax.figure.savefig("sfh_recovery.pdf")

tengri.analysis.plotting.plot_sfh_comparison(model, results, true_params=None, methods=None, figsize=(15, 4))[source]¶

Side-by-side SFH recovery for multiple methods.

BAGPIPES-style multi-panel layout.

Parameters:

model (SEDModel) – Fitted model instance.
results (dict) – Mapping from method name to Posterior (e.g. {"NUTS": post, "VI": post2}).
true_params (dict, optional) – Ground truth parameters for mock recovery plots.
methods (list of str, optional) – Panel order. Defaults to list(results.keys()).
figsize (tuple of float) – Figure size in inches. Default (15, 4).

Returns:

fig (matplotlib Figure)
axes (list of matplotlib Axes)

Examples

from tengri import plot_sfh_comparison

results = {"NUTS": posterior_nuts, "VI": posterior_vi}
fig, axes = plot_sfh_comparison(model, results, true_params=true_params)
fig.savefig("sfh_comparison.pdf")

SED Plots¶

tengri.analysis.plotting.plot_sed_fit(wave_eff, flux_obs, noise, flux_true=None, posterior_draws=None, ax=None, band_names=None, show_residuals=True)[source]¶

Plot observed photometry with model fit — Prospector style.

Parameters:

wave_eff (array_like, shape (n_bands,)) – Effective filter wavelengths. [Angstrom]
flux_obs (array_like, shape (n_bands,)) – Observed photometric fluxes. [erg/s/cm²/Hz]
noise (array_like, shape (n_bands,)) – 1-sigma photometric uncertainties. [erg/s/cm²/Hz]
flux_true (array_like, shape (n_bands,), optional) – True (noiseless) fluxes for mock recovery plots.
posterior_draws (array_like, shape (n_draws, n_bands), optional) – Posterior predictive flux draws for uncertainty shading.
ax (matplotlib Axes, optional) – Axes to plot on. Creates new figure if None.
band_names (list of str, optional) – Filter names for axis labels.
show_residuals (bool) – If True, adds a residual panel below the main SED panel. Default True.

Returns:

fig

Return type:

matplotlib Figure

Examples

from tengri import plot_sed_fit
import numpy as np

wave = np.array([4800.0, 6200.0, 7700.0, 9000.0])
flux = np.array([1.2e-18, 1.8e-18, 2.1e-18, 1.9e-18])
noise = flux * 0.05
fig = plot_sed_fit(wave, flux, noise, show_residuals=False)

tengri.analysis.plotting.plot_spectrum_fit(wave_obs, spec_obs, noise, spec_true=None, spec_draws=None, features=None, z=0.1)[source]¶

Plot spectroscopic fit with residuals — CIGALE style.

Parameters:

wave_obs (array_like, shape (n_pix,)) – Observed-frame wavelength grid. [Angstrom]
spec_obs (array_like, shape (n_pix,)) – Observed spectrum flux density. [erg/s/cm²/Hz]
noise (array_like, shape (n_pix,)) – 1-sigma per-pixel uncertainties. [erg/s/cm²/Hz]
spec_true (array_like, shape (n_pix,), optional) – True (noiseless) spectrum for mock recovery plots.
spec_draws (array_like, shape (n_draws, n_pix), optional) – Posterior predictive spectrum draws for uncertainty shading.
features (dict, optional) – Rest-frame line/feature wavelengths to annotate, e.g. {"Hα": 6563.0}.
z (float) – Redshift for shifting feature wavelengths to observed frame. Default 0.1.

Returns:

fig

Return type:

matplotlib Figure

Examples

from tengri import plot_spectrum_fit

fig = plot_spectrum_fit(
    wave_obs,
    spec_obs,
    noise,
    spec_draws=posterior_spec_draws,
    features={"Hα": 6563.0, "Hβ": 4861.0},
    z=0.5,
)
fig.savefig("spectrum_fit.pdf")

Corner Plots¶

tengri.analysis.plotting.safe_corner(posterior, **kwargs)[source]¶

Wrapper around Posterior.plot_corner that handles degenerate posteriors.

Parameters:

posterior (Posterior) – Fitted posterior object with a plot_corner method.
**kwargs – Additional keyword arguments passed to posterior.plot_corner.

Returns:

fig – Corner plot figure, or None if generation failed.

Return type:

matplotlib.figure.Figure or None

Examples

from tengri import safe_corner

fig = safe_corner(posterior, params=["sfh_dpl_alpha", "dust_tau_bc"])
if fig is not None:
    fig.savefig("corner.pdf")

tengri.analysis.plotting.plot_corner_comparison(posteriors, labels, colors=None, truths=None, params=None)[source]¶

Overlay multiple posteriors on a single corner plot.

Parameters:

posteriors (list of Posterior) – Posterior objects to overlay.
labels (list of str) – Legend labels for each posterior (e.g. ["NUTS", "VI", "MAP"]).
colors (list of str, optional) – Colors for each posterior. Defaults to sampler palette.
truths (dict, optional) – True parameter values for recovery plots.
params (list of str, optional) – Parameter names to include. Defaults to all free parameters.

Returns:

fig – Combined corner plot, or None if all posteriors failed.

Return type:

matplotlib Figure or None

Examples

from tengri import plot_corner_comparison

fig = plot_corner_comparison(
    [post_nuts, post_vi],
    labels=["NUTS", "VI"],
    truths=true_params,
    params=["sfh_dpl_alpha", "dust_tau_bc"],
)
if fig is not None:
    fig.savefig("corner_comparison.pdf")

Diagnostics¶

tengri.analysis.plotting.diagnostics_table(results, names=None)[source]¶

Print a formatted diagnostics comparison table.

Parameters:

results (dict) – Mapping from method name to Posterior (e.g. from Fitter.run()).
names (list of str, optional) – Display order. Defaults to list(results.keys()).

Return type:

None

Examples

from tengri import diagnostics_table

results = {"NUTS": posterior_nuts, "VI": posterior_vi}
diagnostics_table(results)
# Method          Wall time   ESS (min)   ESS (med)   Accept %
# ----------------------------------------------------------------
# NUTS               42.3s         812         934      82.4%
# VI                  3.1s           —           —         —

Constants¶

tengri.analysis.plotting.COLORS = {'data': '#333333', 'g': '#008837', 'geovi': '#ff7f0e', 'i': '#fc8d59', 'map': '#888888', 'mcmc_nuts': '#2ca02c', 'mcmc_raytrace': '#1f77b4', 'mgvi': '#9467bd', 'model': '#d62728', 'nuts': '#2ca02c', 'r': '#d73027', 'rt': '#1f77b4', 'seq': ['#d4d4d4', '#a8a8a8', '#1f77b4', '#2ca02c', '#d62728'], 'sfh_full': '#ff7f0e', 'sfh_gp': '#2ca02c', 'sfh_mean': '#1f77b4', 'truth': '#1a1a1a', 'u': '#7b3294', 'vi': '#ff7f0e', 'vi_linear': '#9467bd', 'z': '#4575b4'}¶

dict() -> new empty dictionary dict(mapping) -> new dictionary initialized from a mapping object’s

(key, value) pairs

dict(iterable) -> new dictionary initialized as if via:: d = {} for k, v in iterable:

d[k] = v
dict(**kwargs) -> new dictionary initialized with the name=value pairs: in the keyword argument list. For example: dict(one=1, two=2)

tengri.analysis.plotting.SDSS_WAVE_EFF = array([3551, 4686, 6166, 7480, 8932])¶

ndarray(shape, dtype=None, buffer=None, offset=0, strides=None, order=None) –

ndarray(shape, dtype=float, buffer=None, offset=0, strides=None, order=None)

An array object represents a multidimensional, homogeneous array of fixed-size items. An associated data-type object describes the format of each element in the array (its byte-order, how many bytes it occupies in memory, whether it is an integer, a floating point number, or something else, etc.)

Arrays should be constructed using array, zeros or empty (refer to the See Also section below). The parameters given here refer to a low-level method (ndarray(…)) for instantiating an array.

For more information, refer to the numpy module and examine the methods and attributes of an array.

Parameters:

below) ((for the __new__ method; see Notes)
shape (tuple of ints) – Shape of created array.
dtype (data-type, optional) – Any object that can be interpreted as a numpy data type. Default is numpy.float64.
buffer (object exposing buffer interface, optional) – Used to fill the array with data.
offset (int, optional) – Offset of array data in buffer.
strides (tuple of ints, optional) – Strides of data in memory.
order ({'C', 'F'}, optional) – Row-major (C-style) or column-major (Fortran-style) order.

tengri.analysis.plotting.T¶

Transpose of the array.

Type:: ndarray

tengri.analysis.plotting.data¶

The array’s elements, in memory.

Type:: buffer

tengri.analysis.plotting.dtype¶

Describes the format of the elements in the array.

Type:: dtype object

tengri.analysis.plotting.flags¶

Dictionary containing information related to memory use, e.g., ‘C_CONTIGUOUS’, ‘OWNDATA’, ‘WRITEABLE’, etc.

Type:: dict

tengri.analysis.plotting.flat¶

Flattened version of the array as an iterator. The iterator allows assignments, e.g., x.flat = 3 (See ndarray.flat for assignment examples; TODO).

Type:: numpy.flatiter object

tengri.analysis.plotting.imag¶

Imaginary part of the array.

Type:: ndarray

tengri.analysis.plotting.real¶

Real part of the array.

Type:: ndarray

tengri.analysis.plotting.size¶

Number of elements in the array.

Type:: int

tengri.analysis.plotting.itemsize¶

The memory use of each array element in bytes.

Type:: int

tengri.analysis.plotting.nbytes¶

The total number of bytes required to store the array data, i.e., itemsize * size.

Type:: int

tengri.analysis.plotting.ndim¶

The array’s number of dimensions.

Type:: int

tengri.analysis.plotting.shape¶

Shape of the array.

Type:: tuple of ints

tengri.analysis.plotting.strides¶

The step-size required to move from one element to the next in memory. For example, a contiguous (3, 4) array of type int16 in C-order has strides (8, 2). This implies that to move from element to element in memory requires jumps of 2 bytes. To move from row-to-row, one needs to jump 8 bytes at a time (2 * 4).

Type:: tuple of ints

tengri.analysis.plotting.ctypes¶

Class containing properties of the array needed for interaction with ctypes.

Type:: ctypes object

tengri.analysis.plotting.base¶

If the array is a view into another array, that array is its base (unless that array is also a view). The base array is where the array data is actually stored.

Type:: ndarray