Plotter

Generally accessible via:

from chainconsumer import ChainConsumer
c = ChainConsumer()
...
c.plotter

chainconsumer.plotter.Plotter

Source code in src/chainconsumer/plotter.py

class Plotter:
    def __init__(self, parent: "ChainConsumer") -> None:
        self.parent: ChainConsumer = parent
        self._config: PlotConfig | None = None
        self._default_config = PlotConfig()

        self.usetex_old = matplotlib.rcParams["text.usetex"]
        self.serif_old = matplotlib.rcParams["font.family"]

    def set_config(self, config: PlotConfig) -> None:
        """Sets the plot config to the chosen `PlotConfig` model.

        Args:
            config: The config to use

        """
        self._config = config

    @property
    def config(self) -> PlotConfig:
        if self._config is None:
            return self._default_config
        return self._config

    def plot(
        self,
        chains: list[ChainName | Chain] | None = None,
        columns: list[ColumnName] | None = None,
        filename: list[str | Path] | str | Path | None = None,
        figsize: FigSize | float | int | tuple[float, float] = FigSize.GROW,
    ) -> Figure:  # pragma: no cover
        """Plot the chain!

        Args:
            chains:
                Used to specify which chain to show if more than one chain is loaded in.
                Can be an integer, specifying the
                chain index, or a str, specifying the chain name.
            columns:
                If set, only creates a plot for those specific parameters (if list). If an
                integer is given, only plots the fist so many parameters.
            filename:
                If set, saves the figure to this location
            figsize:
                The figure size to generate. Accepts a regular two tuple of size in inches,
                or one of several key words. The default value of ``COLUMN`` creates a figure
                of appropriate size of insertion into an A4 LaTeX document in two-column mode.
                ``PAGE`` creates a full page width figure. ``GROW`` creates an image that
                scales with parameters (1.5 inches per parameter). String arguments are not
                case sensitive. If you pass a float, it will scale the default ``GROW`` by
                that amount, so ``2.0`` would result in a plot 3 inches per parameter.

        Returns:
            the matplotlib figure

        """
        base = self._sanitise(
            chains, columns, self.config.extents, blind=self.config.blind, log_scales=self.config.log_scales
        )

        show_legend = self.config.show_legend
        if show_legend is None:
            show_legend = len(base.chains) > 1

        num_cax = len(set([chain.color_param for chain in base.chains if chain.color_param is not None]))
        fig_size = FigSize.get_size(figsize, len(base.columns), num_cax > 0)
        plot_hists = self.config.plot_hists
        flip = len(base.columns) == 2 and plot_hists and self.config.flip
        fig, axes, params_x, params_y = self._get_triangle_figure(base, figsize=fig_size)

        axl = axes.ravel().tolist()
        summarise = self.config.summarise and len(base.chains) == 1

        paths_for_cbar: dict[ColumnName, PathCollection] = {}
        for i, p1 in enumerate(params_x):
            for j, p2 in enumerate(params_y):
                if i < j:
                    continue
                ax: Axes = axes[i, j]
                do_flip = flip and i == len(params_x) - 1

                # Plot the histograms
                if plot_hists and i == j:
                    if do_flip:
                        plot_truths(ax, self.parent._truths, py=p1)
                    else:
                        plot_truths(ax, self.parent._truths, px=p1)
                    max_val = None

                    # Plot each chain
                    for chain in base.chains:
                        if not chain.plot_contour or p1 not in chain.samples:
                            continue

                        do_summary = summarise and p1 not in base.blind
                        max_hist_val = self._plot_bars(ax, p1, chain, flip=do_flip, summary=do_summary)

                        if max_val is None or max_hist_val > max_val:
                            max_val = max_hist_val

                    if max_val is not None:
                        if do_flip:
                            ax.set_xlim(0, 1.1 * max_val)
                        else:
                            ax.set_ylim(0, 1.1 * max_val)

                else:
                    paths_for_cbar |= plot_surface(ax, base.chains, p2, p1, self.config)
                    plot_truths(ax, self.parent._truths, px=p2, py=p1)

        # Create all the colorbars we need
        if paths_for_cbar:
            aspect = fig_size[1] / 0.15
            fraction = 0.85 / fig_size[0]
            for column, path in paths_for_cbar.items():
                cbar = fig.colorbar(path, ax=axl, aspect=aspect, pad=0.03, fraction=fraction, drawedges=False)
                label = self.config.get_label(column)
                if label == "weight":
                    label = "Weights"
                elif label == "log_weight":
                    label = "log(Weights)"
                elif label == "posterior":
                    label = "log(Posterior)"
                cbar.set_label(label, fontsize=self.config.label_font_size)
                if cbar.solids is not None:
                    cbar.solids.set(alpha=1)

        legend_location = self.config.legend_location
        if legend_location is None:
            legend_location = (0, -1) if not flip or len(base.columns) > 2 else (-1, 0)
        legend_outside = legend_location[0] >= legend_location[1]

        if show_legend:
            ax = axes[legend_location[0], legend_location[1]]
            legend_kwargs = self.config.legend_kwargs_final.copy()
            if "markerfirst" not in legend_kwargs:
                legend_kwargs["markerfirst"] = legend_outside or not self.config.legend_artists

            chains_to_show_on_legend = [c for c in base.chains if c.show_label_in_legend]
            artists = get_artists_from_chains(chains_to_show_on_legend)
            leg = ax.legend(handles=artists, **legend_kwargs)
            if self.config.legend_color_text:
                for text, chain in zip(leg.get_texts(), chains_to_show_on_legend):
                    text.set_fontweight("medium")
                    text.set_color(colors.format(chain.color))
        fig.canvas.draw()
        for ax in axes[-1, :]:
            offset = ax.get_xaxis().get_offset_text()
            ax.set_xlabel("{} {}".format(ax.get_xlabel(), f"[{offset.get_text()}]" if offset.get_text() else ""))
            offset.set_visible(False)
        for ax in axes[:, 0]:
            offset = ax.get_yaxis().get_offset_text()
            ax.set_ylabel("{} {}".format(ax.get_ylabel(), f"[{offset.get_text()}]" if offset.get_text() else ""))
            offset.set_visible(False)

        if self.config.watermark is not None:
            ax_watermark = axes[-1, 0] if flip and len(base.columns) == 2 else None
            add_watermark(fig, ax_watermark, fig_size, self.config)

        self._save_fig(fig, filename, dpi=self.config.dpi)

        return fig

    def _save_fig(self, fig: Figure, filename: list[str | Path] | str | Path | None = None, dpi: int = 300) -> None:
        if filename is not None:
            if not isinstance(filename, list):
                filename = [filename]
            for f in filename:
                fig.savefig(f, bbox_inches="tight", dpi=dpi, transparent=True, pad_inches=0.05)

    def plot_walks(
        self,
        chains: list[ChainName | Chain] | None = None,
        columns: list[ColumnName] | None = None,
        filename: list[str | Path] | str | Path | None = None,
        figsize: float | tuple[float, float] | None = None,
        convolve: int | None = None,
        plot_weights: bool = True,
        plot_posterior: bool = True,
        log_weight: bool = False,
    ) -> Figure:  # pragma: no cover
        """Plots the chain walk; the parameter values as a function of step index.

        This plot is more for a sanity or consistency check than for use with final results.
        Plotting this before plotting with :func:`plot` allows you to quickly see if the
        chains are well behaved, or if certain parameters are suspect
        or require a greater burn in period.

        The desired outcome is to see an unchanging distribution along the x-axis of the plot.
        If there are obvious tails or features in the parameters, you probably want
        to investigate.

        Args:
            chains:
                Used to specify which chain to show if more than one chain is loaded in.
                Can be an integer, specifying the
                chain index, or a str, specifying the chain name.
            columns:
                If set, only creates a plot for those specific parameters (if list). If an
                integer is given, only plots the fist so many parameters.
            filename:
                If set, saves the figure to this location
            figsize:
                Scale horizontal and vertical figure size.
            col_wrap:
                How many columns to plot before wrapping.
            convolve:
                If set, overplots a smoothed version of the steps using ``convolve`` as
                the width of the smoothing filter.
            plot_weights:
                If true, plots the weight if they are available
            plot_posterior:
                If true, plots the log posterior if they are available
            log_weight:
                Whether to display weights in log space or not. If None, the value is
                inferred by the mean weights of the plotted chains.

        Returns:
            the matplotlib figure created

        """

        base = self._sanitise(
            chains,
            columns,
            self.config.extents,
            blind=self.config.blind,
            log_scales=self.config.log_scales,
        )

        n = len(base.columns)
        extra = 0

        plot_posterior = plot_posterior and any([c.log_posterior is not None for c in base.chains])
        if plot_weights:
            extra += 1
        if plot_posterior:
            extra += 1

        if figsize is None:
            fig_size = (8, 0.75 + (n + extra))
        elif isinstance(figsize, float | int):
            fig_size = (figsize, figsize)
        else:
            fig_size = figsize

        fig, axes = plt.subplots(figsize=fig_size, nrows=n + extra, squeeze=False, sharex=True)
        max_points = 100000
        for i, axes_row in enumerate(axes):
            ax = axes_row[0]
            if i >= extra:
                p = base.columns[i - extra]
                for chain in base.chains:
                    if p in chain.data_columns:
                        chain_row = chain.get_data(p)
                        if len(chain_row) > max_points:
                            chain_row = chain_row[:: int(len(chain_row) / max_points)]
                        log = p in base.log_scales
                        self._plot_walk(
                            ax,
                            p,
                            chain_row,
                            extents=base.extents.get(p),
                            convolve=convolve,
                            color=colors.format(chain.color),
                            log_scale=log,
                        )
                for truth in self.parent._truths:
                    if p in truth.location:
                        self._plot_walk_truth(ax, truth, p)

                if p in base.blind:
                    ax.set_yticks([])
            else:  # noqa: PLR5501
                if i == 0 and plot_posterior:
                    for chain in base.chains:
                        if chain.log_posterior is not None:
                            posterior = chain.log_posterior - chain.log_posterior.max()
                            if len(posterior) > max_points:
                                posterior = posterior[:: int(len(posterior) / max_points)]

                            self._plot_walk(
                                ax,
                                r"$\log(P)$",
                                posterior,
                                convolve=convolve,
                                color=colors.format(chain.color),
                            )
                else:
                    label = r"$\log_{10}$Weight" if log_weight else "Weight"

                    for chain in base.chains:
                        if chain.weights is not None:
                            weights = chain.weights
                            if len(weights) > max_points:
                                weights = weights[:: int(len(weights) / max_points)]
                            self._plot_walk(
                                ax,
                                label,
                                np.log10(weights) if log_weight else weights,  # type: ignore
                                convolve=convolve,
                                color=colors.format(chain.color),
                            )

        add_watermark(fig, None, fig_size, self.config, size_scale=0.8)
        self._save_fig(fig, filename, dpi=self.config.dpi)

        return fig

    def plot_distributions(
        self,
        chains: list[ChainName | Chain] | None = None,
        columns: list[ColumnName] | None = None,
        filename: list[str | Path] | str | Path | None = None,
        col_wrap: int = 4,
        figsize: float | tuple[float, float] | None = None,
    ) -> Figure:  # pragma: no cover
        """Plots the 1D parameter distributions for verification purposes.

        This plot is more for a sanity or consistency check than for use with final results.
        Plotting this before plotting with :func:`plot` allows you to quickly see if the
        chains give well behaved distributions, or if certain parameters are suspect
        or require a greater burn in period.

        Args:
            chains:
                Used to specify which chain to show if more than one chain is loaded in.
                Can be an integer, specifying the
                chain index, or a str, specifying the chain name.
            columns:
                If set, only creates a plot for those specific parameters (if list). If an
                integer is given, only plots the fist so many parameters.
            filename:
                If set, saves the figure to this location
            figsize:
                Scale horizontal and vertical figure size.
            col_wrap:
                How many columns to plot before wrapping.

        Returns:
            the matplotlib figure created

        """
        base = self._sanitise(
            chains,
            columns,
            self.config.extents,
            blind=self.config.blind,
            log_scales=self.config.log_scales,
        )

        n = len(base.columns)
        num_cols = min(n, col_wrap)
        num_rows = int(np.ceil(1.0 * n / col_wrap))

        if figsize is None:
            figsize = 1.0
        if isinstance(figsize, float | int):
            figsize_float = figsize
            figsize = (num_cols * 2.5 * figsize, num_rows * 2.5 * figsize)
        else:
            figsize_float = 1.0

        summary = self.config.summarise and len(base.chains) == 1
        hspace = (0.8 if summary else 0.5) / figsize_float
        fig, axes = plt.subplots(nrows=num_rows, ncols=num_cols, figsize=figsize, squeeze=False)
        fig.subplots_adjust(left=0.1, right=0.95, top=0.95, bottom=0.1, wspace=0.05, hspace=hspace)

        formatter = ScalarFormatter(useOffset=False)
        formatter.set_powerlimits((-3, 4))

        for i, ax in enumerate(axes.flatten()):
            if i >= n:
                ax.set_axis_off()
                continue
            p = base.columns[i]

            ax.set_yticks([])
            if p in base.log_scales:
                ax.set_xscale("log")
            if p in base.blind:
                ax.set_xticks([])
            else:
                if self.config.diagonal_tick_labels:
                    _ = [label.set_rotation(45) for label in ax.get_xticklabels()]
                _ = [label.set_fontsize(self.config.tick_font_size) for label in ax.get_xticklabels()]

                if p in base.log_scales:
                    ax.xaxis.set_major_locator(LogLocator(numticks=self.config.max_ticks))
                else:
                    ax.xaxis.set_major_locator(MaxNLocator(self.config.max_ticks, prune="lower"))
                    ax.xaxis.set_major_formatter(formatter)
            ax.set_xlim(base.extents.get(p) or self._get_parameter_extents(p, base.chains))

            max_val = -np.inf
            for chain in base.chains:
                if not chain.plot_contour:
                    continue
                if p in chain.plotting_columns:
                    param_summary = summary and p not in base.blind
                    m = self._plot_bars(ax, p, chain, summary=param_summary)
                    if max_val is None or m > max_val:
                        max_val = m
            plot_truths(ax, self.parent._truths, py=p)
            ax.set_ylim(0, 1.1 * max_val)
            ax.set_xlabel(p, fontsize=self.config.label_font_size)

        add_watermark(fig, None, figsize, self.config, size_scale=0.8)
        self._save_fig(fig, filename, dpi=self.config.dpi)
        return fig

    def plot_summary(
        self,
        chains: list[ChainName | Chain] | None = None,
        columns: list[ColumnName] | None = None,
        filename: list[str | Path] | str | Path | None = None,
        figsize: float = 1.0,
        errorbar: bool = False,
        extra_parameter_spacing: float = 1.0,
        vertical_spacing_ratio: float = 1.0,
    ) -> Figure:  # pragma: no cover
        """Plots parameter summaries

        This plot is more for a sanity or consistency check than for use with final results.
        Plotting this before plotting with :func:`plot` allows you to quickly see if the
        chains give well behaved distributions, or if certain parameters are suspect
        or require a greater burn in period.

        Args:
            chains:
                Used to specify which chain to show if more than one chain is loaded in.
                Can be an integer, specifying the
                chain index, or a str, specifying the chain name.
            columns:
                If set, only creates a plot for those specific parameters (if list). If an
                integer is given, only plots the fist so many parameters.
            filename:
                If set, saves the figure to this location
            figsize:
                Scale horizontal and vertical figure size.
            errorbar:
                Whether to onle plot an error bar, instead of the marginalised distribution.
            include_truth_chain:
                If you specify another chain as the truth chain, determine if it should still
                be plotted.
            extra_parameter_spacing:
                Increase horizontal space for parameter values
            vertical_spacing_ratio:
                Increase vertical space for each model
        Returns:
            the matplotlib figure created

        """
        wide_extents = not errorbar
        base = self._sanitise(
            chains,
            columns,
            self.config.extents,
            blind=self.config.blind,
            log_scales=self.config.log_scales,
            wide_extents=wide_extents,
        )

        # We have a bit of fun to go from chain names to the width of the
        # subplot used to display said names
        max_param = self._get_size_of_texts(base.columns)
        fid_dpi = 65  # Seriously I have no idea what value this should be
        param_width = extra_parameter_spacing + max(0.5, max_param / fid_dpi)
        max_model_name = self._get_size_of_texts([chain.name for chain in base.chains])
        model_width = 0.25 + (max_model_name / fid_dpi)
        gridspec_kw = {
            "width_ratios": [model_width] + [param_width] * len(base.columns),
            "height_ratios": [1] * len(base.chains),
        }
        ncols = 1 + len(base.columns)
        top_spacing = 0.3
        bottom_spacing = 0.2
        row_height = (0.5 if errorbar else 0.8) * vertical_spacing_ratio
        width = param_width * len(base.columns) + model_width
        height = top_spacing + bottom_spacing + row_height * len(base.chains)
        top_ratio = 1 - (top_spacing / height)
        bottom_ratio = bottom_spacing / height

        fig_size = (width * figsize, height * figsize)
        fig, axes = plt.subplots(
            nrows=len(base.chains), ncols=ncols, figsize=fig_size, squeeze=False, gridspec_kw=gridspec_kw
        )
        fig.subplots_adjust(left=0.05, right=0.95, top=top_ratio, bottom=bottom_ratio, wspace=0.0, hspace=0.0)
        label_font_size = self.config.label_font_size
        legend_color_text = self.config.legend_color_text

        max_vals: dict[ColumnName, float] = {}
        num_chains = len(base.chains)
        for i, axes_row in enumerate(axes):
            chain = base.chains[i]
            colour = colors.format(chain.color)

            # First one put name of model
            ax_first = axes_row[0]
            ax_first.set_axis_off()
            text_colour = "k" if not legend_color_text else colour
            ax_first.text(
                0,
                0.5,
                chain.name,
                transform=ax_first.transAxes,
                fontsize=label_font_size,
                verticalalignment="center",
                color=text_colour,
                weight="medium",
            )
            axes_for_summaries = axes_row[1:]

            for ax, p in zip(axes_for_summaries, base.columns):
                # Set up the frames
                if i > 0:
                    ax.spines["top"].set_visible(False)
                if i < (num_chains - 1):
                    ax.spines["bottom"].set_visible(False)
                if i < (num_chains - 1) or p in base.blind:
                    ax.set_xticks([])
                ax.set_yticks([])
                ax.set_xlim(base.extents[p])
                if p in base.log_scales:
                    ax.set_xscale("log")

                # Put title in
                if i == 0:
                    ax.set_title(self.config.get_label(p), fontsize=label_font_size)

                # Add truth values
                for truth in self.parent._truths:
                    truth_value = truth.location.get(p)
                    if truth_value is not None:
                        ax.axvline(truth_value, **truth._kwargs)

                # Skip if this chain doesnt have the parameter
                if p not in chain.data_columns:
                    continue

                # Plot the good stuff
                if errorbar:
                    fv = self.parent.analysis.get_parameter_summary(chain, p)
                    if fv is None or fv.all_none:
                        continue
                    if fv.lower is not None and fv.upper is not None:
                        diff = np.abs(np.diff(fv.array))
                        ax.errorbar([fv.center], 0, xerr=[[diff[0]], [diff[1]]], fmt="o", color=colour)
                else:
                    m = self._plot_bars(ax, p, chain)
                    if max_vals.get(p) is None or m > max_vals[p]:
                        max_vals[p] = m

        for i, axes_row in enumerate(axes):
            for ax, p in zip(axes_row, base.columns):
                if not errorbar:
                    ax.set_ylim(0, 1.1 * max_vals[p])

        add_watermark(fig, None, fig_size, self.config, size_scale=0.8)
        self._save_fig(fig, filename, dpi=self.config.dpi)

        return fig

    def _get_size_of_texts(self, texts: list[str]) -> float:  # pragma: no cover
        usetex = self.config.usetex
        size = self.config.label_font_size
        widths = [TextPath((0, 0), text, usetex=usetex, size=size).get_extents().width for text in texts]
        return max(widths)

    def _sanitise_columns(self, columns: list[ColumnName] | None, chains: list[Chain]) -> list[ColumnName]:
        if columns is None:
            res = []  # Doing it without set to preserve order
            for chain in chains:
                for column in chain.plotting_columns:
                    if column not in res:
                        res.append(column)
            return res
        return columns

    def _sanitise_logscale(self, log_scales: list[ColumnName] | None) -> list[ColumnName]:
        # We could at some point determine if something should be a log scale by analyising
        # its distribution, but for now assume its all linear
        if log_scales is None:
            return []
        return log_scales

    def _sanitise_blinds(self, blind: bool | list[ColumnName] | None, columns: list[ColumnName]) -> list[ColumnName]:
        if blind is None or blind is False:
            return []
        elif blind is True:
            return columns
        return blind

    def _sanitise(
        self,
        chains: list[ChainName | Chain] | None,
        columns: list[ColumnName] | None,
        extents: dict[str, tuple[float, float]] | None,
        blind: bool | list[ColumnName] | None = None,
        log_scales: list[ColumnName] | None = None,
        wide_extents: bool = True,
    ) -> PlottingBase:
        final_chains = self._sanitise_chains(chains)
        final_columns = self._sanitise_columns(columns, final_chains)
        extents = self._get_custom_extents(final_columns, final_chains, extents, wide_extents=wide_extents)
        self.set_rc_params()

        return PlottingBase(
            chains=final_chains,
            columns=final_columns,
            extents=extents,
            log_scales=self._sanitise_logscale(log_scales),
            blind=self._sanitise_blinds(blind, final_columns),
        )

    def set_rc_params(self) -> None:
        if self.config.usetex:
            plt.rc("text", usetex=True)
        else:
            plt.rc("text", usetex=False)
        if self.config.serif:
            plt.rc("font", family="serif")
        else:
            plt.rc("font", family="sans-serif")

    def restore_rc_params(self):
        """Restores the matplotlib rc parameters modified by usetex and serif.

        Unfortunately this cannot be automated because you cannot invoke it whilst you have
        an active figure object or matplotlib will destroy you. So do all your plotting, close
        the plots, and then you can call this.
        """
        plt.rc("text", usetex=self.usetex_old)
        plt.rc("font", family=self.serif_old)

    def _get_custom_extents(
        self,
        columns: list[ColumnName],
        chains: list[Chain],
        initial_extents: dict[ColumnName, tuple[float, float]] | None,
        wide_extents: bool = True,
    ) -> dict[ColumnName, tuple[float, float]]:  # pragma: no cover
        if initial_extents is None:
            initial_extents = {}
        extents = {} | initial_extents
        for p in columns:
            if p not in initial_extents:
                extents[p] = self._get_parameter_extents(p, chains, wide_extents=wide_extents)
        return extents

    def _get_triangle_figure(
        self, base: PlottingBase, figsize: tuple[float, float]
    ) -> tuple[Figure, np.ndarray, list[ColumnName], list[ColumnName]]:
        n = len(base.columns)
        if not self.config.plot_hists:
            n -= 1

        spacing = self.config.spacing
        if spacing is None:
            spacing = 1.0 if n < 6 else 0.0

        gridspec_kw = {}
        if n == 2 and self.config.plot_hists and self.config.flip:
            gridspec_kw = {"width_ratios": [3, 1], "height_ratios": [1, 3]}

        fig, axes = plt.subplots(n, n, figsize=figsize, squeeze=False, gridspec_kw=gridspec_kw)
        min_left_for_axes = min(max(0.85 / figsize[0], 0.1), 0.3)
        min_bottom_for_axes = min(max(0.85 / figsize[1], 0.1), 0.3)
        fig.subplots_adjust(
            left=min_left_for_axes,
            right=0.95,
            top=0.9,
            bottom=min_bottom_for_axes,
            wspace=0.05 * spacing,
            hspace=0.05 * spacing,
        )

        if self.config.plot_hists:
            params_x = base.columns
            params_y = base.columns
        else:
            params_x = base.columns[1:]
            params_y = base.columns[:-1]
        for i, p1 in enumerate(params_x):
            for j, p2 in enumerate(params_y):
                ax = axes[i, j]
                formatter_x = ScalarFormatter(useOffset=True)
                formatter_x.set_powerlimits((-3, 4))
                formatter_y = ScalarFormatter(useOffset=True)
                formatter_y.set_powerlimits((-3, 4))

                display_x_ticks = False
                display_y_ticks = False
                if i < j:
                    ax.set_frame_on(False)
                    ax.set_xticks([])
                    ax.set_yticks([])
                else:
                    logx = False
                    logy = False
                    if p1 == p2:
                        if p1 in base.log_scales:
                            if self.config.flip and j == n - 1:
                                ax.set_yscale("log")
                                logy = True
                            else:
                                ax.set_xscale("log")
                                logx = True
                    else:
                        if p1 in base.log_scales:
                            ax.set_yscale("log")
                            logy = True
                        if p2 in base.log_scales:
                            ax.set_xscale("log")
                            logx = True
                    if i != n - 1 or (self.config.flip and j == n - 1):
                        ax.set_xticks([])
                    else:
                        if p2 in base.blind:
                            ax.set_xticks([])
                        else:
                            display_x_ticks = True
                        if isinstance(p2, str):
                            ax.set_xlabel(self.config.get_label(p2), fontsize=self.config.label_font_size)
                    if j != 0 or (self.config.plot_hists and i == 0):
                        ax.set_yticks([])
                    else:
                        if p1 in base.blind:
                            ax.set_yticks([])
                        else:
                            display_y_ticks = True
                        if isinstance(p1, str):
                            ax.set_ylabel(self.config.get_label(p1), fontsize=self.config.label_font_size)
                    if display_x_ticks:
                        if self.config.diagonal_tick_labels:
                            _ = [label.set_rotation(45) for label in ax.get_xticklabels()]
                        _ = [label.set_fontsize(self.config.tick_font_size) for label in ax.get_xticklabels()]
                        if not logx:
                            ax.xaxis.set_major_locator(MaxNLocator(self.config.max_ticks, prune="lower"))
                            ax.xaxis.set_major_formatter(formatter_x)
                        else:
                            ax.xaxis.set_major_locator(LogLocator(numticks=self.config.max_ticks))
                    else:
                        ax.set_xticks([])
                    if display_y_ticks:
                        if self.config.diagonal_tick_labels:
                            _ = [label.set_rotation(45) for label in ax.get_yticklabels()]
                        _ = [label.set_fontsize(self.config.tick_font_size) for label in ax.get_yticklabels()]
                        if not logy:
                            ax.yaxis.set_major_locator(MaxNLocator(self.config.max_ticks, prune="lower"))
                            ax.yaxis.set_major_formatter(formatter_y)
                        else:
                            ax.yaxis.set_major_locator(LogLocator(numticks=self.config.max_ticks))
                    else:
                        ax.set_yticks([])
                    if (i != j or not self.config.plot_hists) or (self.config.flip and i == 1):
                        ax.set_ylim(base.extents[p1])
                    ax.set_xlim(base.extents[p2])

        return fig, axes, params_x, params_y

    def _get_parameter_extents(
        self, column: ColumnName, chains: list[Chain], wide_extents: bool = True
    ) -> tuple[float, float]:
        min_val, max_val = np.inf, -np.inf
        for chain in chains:
            if column not in chain.samples:
                continue  # pragma: no cover

            data = chain.get_data(column)
            min_prop, max_prop = np.inf, -np.inf
            if chain.plot_contour or chain.plot_cloud:
                if chain.grid:
                    min_prop = data.min()
                    max_prop = data.max()
                else:
                    min_prop, max_prop = get_extents(data, chain.weights, plot=True, wide_extents=wide_extents)

            else:
                point = chain.get_max_posterior_point()
                if point is not None and column in point.coordinate:
                    min_prop = point.coordinate[column]
                    max_prop = min_prop

            min_val = min(min_prop, min_val)
            max_val = max(max_prop, max_val)

        return min_val, max_val

    def _sanitise_chains(
        self, chains: list[Chain | ChainName] | dict[ChainName, Chain] | None, include_skip: bool = False
    ) -> list[Chain]:
        overriden_chains = self.parent._get_final_chains()
        final_chains = []
        if isinstance(chains, list):
            final_chains = [overriden_chains[c if isinstance(c, ChainName) else c.name] for c in chains]
        elif isinstance(chains, dict):
            final_chains = [overriden_chains[c.name] for c in chains.values()]
        else:
            final_chains = list(overriden_chains.values())
        return [c for c in final_chains if include_skip or not c.skip]

    def _plot_bars(
        self, ax: Axes, column: str, chain: Chain, flip: bool = False, summary: bool = False
    ) -> float:  # pragma: no cover
        # Get values from config
        data = chain.get_data(column)
        if chain.smooth or chain.kde:
            xs, ys, _ = self.parent.analysis._get_smoothed_histogram(chain, column, pad=True)
            ys *= chain.histogram_relative_height
            if flip:
                ax.plot(ys, xs, color=chain.color, ls=chain.linestyle, lw=chain.linewidth, zorder=chain.zorder)
            else:
                ax.plot(xs, ys, color=chain.color, ls=chain.linestyle, lw=chain.linewidth, zorder=chain.zorder)
        else:
            if chain.grid:
                bins = get_grid_bins(data)
            else:
                bins, _ = get_smoothed_bins(chain.smooth, get_bins(chain), data, chain.weights)
            hist, edges = np.histogram(data, bins=bins, density=True, weights=chain.weights)
            if chain.power is not None:
                hist = hist**chain.power
            edge_center = 0.5 * (edges[:-1] + edges[1:])
            xs, ys = edge_center, hist
            ys *= chain.histogram_relative_height
            ax.hist(
                xs,
                weights=ys,
                bins=bins,  # type: ignore
                histtype="step",
                color=chain.color,  # type: ignore
                orientation="horizontal" if flip else "vertical",
                ls=chain.linestyle,
                lw=chain.linewidth,
                zorder=chain.zorder,
            )
        interp_type = "linear" if chain.smooth else "nearest"
        interpolator = interp1d(xs, ys, kind=interp_type)

        if chain.bar_shade:
            fit_values = self.parent.analysis.get_parameter_summary(chain, column)
            if fit_values is not None:
                lower = fit_values.lower
                upper = fit_values.upper
                if lower is not None and upper is not None:
                    lower = max(lower, xs.min())
                    upper = min(upper, xs.max())
                    x = np.linspace(lower, upper, 1000)  # type: ignore
                    if flip:
                        ax.fill_betweenx(
                            x,
                            np.zeros(x.shape),
                            interpolator(x),
                            color=chain.color,
                            alpha=0.2,
                            zorder=chain.zorder,
                        )
                    else:
                        ax.fill_between(
                            x,
                            np.zeros(x.shape),
                            interpolator(x),
                            color=chain.color,
                            alpha=0.2,
                            zorder=chain.zorder,
                        )
                    if summary:
                        t = self.parent.analysis.get_parameter_text(fit_values)
                        label = self.config.get_label(column)
                        if isinstance(column, str):
                            ax.set_title(
                                r"${} = {}$".format(label.strip("$"), t), fontsize=self.config.summary_font_size
                            )
                        else:
                            ax.set_title(rf"${t}$", fontsize=self.config.summary_font_size)
        return float(ys.max())

    def _plot_walk(
        self,
        ax: Axes,
        column: ColumnName,
        data: pd.Series,
        extents: tuple[float, float] | None = None,
        convolve: int | None = None,
        color: str | None = None,
        log_scale: bool = False,
    ) -> None:  # pragma: no cover
        if extents is not None:
            ax.set_ylim(extents)
        assert convolve is None or isinstance(convolve, int), "Convolve must be an integer pixel window width"
        x = np.arange(data.size)
        ax.set_xlim(0, x[-1])
        ax.set_ylabel(self.config.get_label(column))
        if color is None:
            color = "#0345A1"
        ax.scatter(x, data, c=color, s=2, marker=".", edgecolors="none", alpha=0.5)
        max_ticks = self.config.max_ticks
        if log_scale:
            ax.set_yscale("log")
            ax.yaxis.set_major_locator(LogLocator(numticks=max_ticks))
        else:
            ax.yaxis.set_major_locator(MaxNLocator(max_ticks, prune="lower"))

        if convolve is not None:
            trim = int(0.5 * convolve)
            color2 = colors.scale_colour(color, 0.5)
            filt = np.ones(convolve) / convolve
            filtered = np.convolve(data, filt, mode="same")
            ax.plot(x[trim:-trim], filtered[trim:-trim], color=color2, alpha=1)

    def _plot_walk_truth(self, ax: Axes, truth: Truth, col: str) -> None:
        ax.axhline(truth.location[col], **truth._kwargs)

plot

plot(chains: list[ChainName | Chain] | None = None, columns: list[ColumnName] | None = None, filename: list[str | Path] | str | Path | None = None, figsize: FigSize | float | int | tuple[float, float] = FigSize.GROW) -> Figure

Plot the chain!

Parameters:

Name	Type	Description	Default
chains	list[ChainName | Chain] | None	Used to specify which chain to show if more than one chain is loaded in. Can be an integer, specifying the chain index, or a str, specifying the chain name.	None
columns	list[ColumnName] | None	If set, only creates a plot for those specific parameters (if list). If an integer is given, only plots the fist so many parameters.	None
filename	list[str | Path] | str | Path | None	If set, saves the figure to this location	None
figsize	FigSize | float | int | tuple[float, float]	The figure size to generate. Accepts a regular two tuple of size in inches, or one of several key words. The default value of COLUMN creates a figure of appropriate size of insertion into an A4 LaTeX document in two-column mode. PAGE creates a full page width figure. GROW creates an image that scales with parameters (1.5 inches per parameter). String arguments are not case sensitive. If you pass a float, it will scale the default GROW by that amount, so 2.0 would result in a plot 3 inches per parameter.	GROW

Returns:

Type	Description
Figure	the matplotlib figure

Source code in src/chainconsumer/plotter.py

def plot(
    self,
    chains: list[ChainName | Chain] | None = None,
    columns: list[ColumnName] | None = None,
    filename: list[str | Path] | str | Path | None = None,
    figsize: FigSize | float | int | tuple[float, float] = FigSize.GROW,
) -> Figure:  # pragma: no cover
    """Plot the chain!

    Args:
        chains:
            Used to specify which chain to show if more than one chain is loaded in.
            Can be an integer, specifying the
            chain index, or a str, specifying the chain name.
        columns:
            If set, only creates a plot for those specific parameters (if list). If an
            integer is given, only plots the fist so many parameters.
        filename:
            If set, saves the figure to this location
        figsize:
            The figure size to generate. Accepts a regular two tuple of size in inches,
            or one of several key words. The default value of ``COLUMN`` creates a figure
            of appropriate size of insertion into an A4 LaTeX document in two-column mode.
            ``PAGE`` creates a full page width figure. ``GROW`` creates an image that
            scales with parameters (1.5 inches per parameter). String arguments are not
            case sensitive. If you pass a float, it will scale the default ``GROW`` by
            that amount, so ``2.0`` would result in a plot 3 inches per parameter.

    Returns:
        the matplotlib figure

    """
    base = self._sanitise(
        chains, columns, self.config.extents, blind=self.config.blind, log_scales=self.config.log_scales
    )

    show_legend = self.config.show_legend
    if show_legend is None:
        show_legend = len(base.chains) > 1

    num_cax = len(set([chain.color_param for chain in base.chains if chain.color_param is not None]))
    fig_size = FigSize.get_size(figsize, len(base.columns), num_cax > 0)
    plot_hists = self.config.plot_hists
    flip = len(base.columns) == 2 and plot_hists and self.config.flip
    fig, axes, params_x, params_y = self._get_triangle_figure(base, figsize=fig_size)

    axl = axes.ravel().tolist()
    summarise = self.config.summarise and len(base.chains) == 1

    paths_for_cbar: dict[ColumnName, PathCollection] = {}
    for i, p1 in enumerate(params_x):
        for j, p2 in enumerate(params_y):
            if i < j:
                continue
            ax: Axes = axes[i, j]
            do_flip = flip and i == len(params_x) - 1

            # Plot the histograms
            if plot_hists and i == j:
                if do_flip:
                    plot_truths(ax, self.parent._truths, py=p1)
                else:
                    plot_truths(ax, self.parent._truths, px=p1)
                max_val = None

                # Plot each chain
                for chain in base.chains:
                    if not chain.plot_contour or p1 not in chain.samples:
                        continue

                    do_summary = summarise and p1 not in base.blind
                    max_hist_val = self._plot_bars(ax, p1, chain, flip=do_flip, summary=do_summary)

                    if max_val is None or max_hist_val > max_val:
                        max_val = max_hist_val

                if max_val is not None:
                    if do_flip:
                        ax.set_xlim(0, 1.1 * max_val)
                    else:
                        ax.set_ylim(0, 1.1 * max_val)

            else:
                paths_for_cbar |= plot_surface(ax, base.chains, p2, p1, self.config)
                plot_truths(ax, self.parent._truths, px=p2, py=p1)

    # Create all the colorbars we need
    if paths_for_cbar:
        aspect = fig_size[1] / 0.15
        fraction = 0.85 / fig_size[0]
        for column, path in paths_for_cbar.items():
            cbar = fig.colorbar(path, ax=axl, aspect=aspect, pad=0.03, fraction=fraction, drawedges=False)
            label = self.config.get_label(column)
            if label == "weight":
                label = "Weights"
            elif label == "log_weight":
                label = "log(Weights)"
            elif label == "posterior":
                label = "log(Posterior)"
            cbar.set_label(label, fontsize=self.config.label_font_size)
            if cbar.solids is not None:
                cbar.solids.set(alpha=1)

    legend_location = self.config.legend_location
    if legend_location is None:
        legend_location = (0, -1) if not flip or len(base.columns) > 2 else (-1, 0)
    legend_outside = legend_location[0] >= legend_location[1]

    if show_legend:
        ax = axes[legend_location[0], legend_location[1]]
        legend_kwargs = self.config.legend_kwargs_final.copy()
        if "markerfirst" not in legend_kwargs:
            legend_kwargs["markerfirst"] = legend_outside or not self.config.legend_artists

        chains_to_show_on_legend = [c for c in base.chains if c.show_label_in_legend]
        artists = get_artists_from_chains(chains_to_show_on_legend)
        leg = ax.legend(handles=artists, **legend_kwargs)
        if self.config.legend_color_text:
            for text, chain in zip(leg.get_texts(), chains_to_show_on_legend):
                text.set_fontweight("medium")
                text.set_color(colors.format(chain.color))
    fig.canvas.draw()
    for ax in axes[-1, :]:
        offset = ax.get_xaxis().get_offset_text()
        ax.set_xlabel("{} {}".format(ax.get_xlabel(), f"[{offset.get_text()}]" if offset.get_text() else ""))
        offset.set_visible(False)
    for ax in axes[:, 0]:
        offset = ax.get_yaxis().get_offset_text()
        ax.set_ylabel("{} {}".format(ax.get_ylabel(), f"[{offset.get_text()}]" if offset.get_text() else ""))
        offset.set_visible(False)

    if self.config.watermark is not None:
        ax_watermark = axes[-1, 0] if flip and len(base.columns) == 2 else None
        add_watermark(fig, ax_watermark, fig_size, self.config)

    self._save_fig(fig, filename, dpi=self.config.dpi)

    return fig

set_config

set_config(config: PlotConfig) -> None

Sets the plot config to the chosen PlotConfig model.

Parameters:

Name	Type	Description	Default
config	PlotConfig	The config to use	required

Source code in src/chainconsumer/plotter.py

def set_config(self, config: PlotConfig) -> None:
    """Sets the plot config to the chosen `PlotConfig` model.

    Args:
        config: The config to use

    """
    self._config = config

plot_walks

plot_walks(chains: list[ChainName | Chain] | None = None, columns: list[ColumnName] | None = None, filename: list[str | Path] | str | Path | None = None, figsize: float | tuple[float, float] | None = None, convolve: int | None = None, plot_weights: bool = True, plot_posterior: bool = True, log_weight: bool = False) -> Figure

Plots the chain walk; the parameter values as a function of step index.

This plot is more for a sanity or consistency check than for use with final results. Plotting this before plotting with :func:plot allows you to quickly see if the chains are well behaved, or if certain parameters are suspect or require a greater burn in period.

The desired outcome is to see an unchanging distribution along the x-axis of the plot. If there are obvious tails or features in the parameters, you probably want to investigate.

Parameters:

Name	Type	Description	Default
chains	list[ChainName | Chain] | None	Used to specify which chain to show if more than one chain is loaded in. Can be an integer, specifying the chain index, or a str, specifying the chain name.	None
columns	list[ColumnName] | None	If set, only creates a plot for those specific parameters (if list). If an integer is given, only plots the fist so many parameters.	None
filename	list[str | Path] | str | Path | None	If set, saves the figure to this location	None
figsize	float | tuple[float, float] | None	Scale horizontal and vertical figure size.	None
col_wrap		How many columns to plot before wrapping.	required
convolve	int | None	If set, overplots a smoothed version of the steps using convolve as the width of the smoothing filter.	None
plot_weights	bool	If true, plots the weight if they are available	True
plot_posterior	bool	If true, plots the log posterior if they are available	True
log_weight	bool	Whether to display weights in log space or not. If None, the value is inferred by the mean weights of the plotted chains.	False

Returns:

Type	Description
Figure	the matplotlib figure created

Source code in src/chainconsumer/plotter.py

def plot_walks(
    self,
    chains: list[ChainName | Chain] | None = None,
    columns: list[ColumnName] | None = None,
    filename: list[str | Path] | str | Path | None = None,
    figsize: float | tuple[float, float] | None = None,
    convolve: int | None = None,
    plot_weights: bool = True,
    plot_posterior: bool = True,
    log_weight: bool = False,
) -> Figure:  # pragma: no cover
    """Plots the chain walk; the parameter values as a function of step index.

    This plot is more for a sanity or consistency check than for use with final results.
    Plotting this before plotting with :func:`plot` allows you to quickly see if the
    chains are well behaved, or if certain parameters are suspect
    or require a greater burn in period.

    The desired outcome is to see an unchanging distribution along the x-axis of the plot.
    If there are obvious tails or features in the parameters, you probably want
    to investigate.

    Args:
        chains:
            Used to specify which chain to show if more than one chain is loaded in.
            Can be an integer, specifying the
            chain index, or a str, specifying the chain name.
        columns:
            If set, only creates a plot for those specific parameters (if list). If an
            integer is given, only plots the fist so many parameters.
        filename:
            If set, saves the figure to this location
        figsize:
            Scale horizontal and vertical figure size.
        col_wrap:
            How many columns to plot before wrapping.
        convolve:
            If set, overplots a smoothed version of the steps using ``convolve`` as
            the width of the smoothing filter.
        plot_weights:
            If true, plots the weight if they are available
        plot_posterior:
            If true, plots the log posterior if they are available
        log_weight:
            Whether to display weights in log space or not. If None, the value is
            inferred by the mean weights of the plotted chains.

    Returns:
        the matplotlib figure created

    """

    base = self._sanitise(
        chains,
        columns,
        self.config.extents,
        blind=self.config.blind,
        log_scales=self.config.log_scales,
    )

    n = len(base.columns)
    extra = 0

    plot_posterior = plot_posterior and any([c.log_posterior is not None for c in base.chains])
    if plot_weights:
        extra += 1
    if plot_posterior:
        extra += 1

    if figsize is None:
        fig_size = (8, 0.75 + (n + extra))
    elif isinstance(figsize, float | int):
        fig_size = (figsize, figsize)
    else:
        fig_size = figsize

    fig, axes = plt.subplots(figsize=fig_size, nrows=n + extra, squeeze=False, sharex=True)
    max_points = 100000
    for i, axes_row in enumerate(axes):
        ax = axes_row[0]
        if i >= extra:
            p = base.columns[i - extra]
            for chain in base.chains:
                if p in chain.data_columns:
                    chain_row = chain.get_data(p)
                    if len(chain_row) > max_points:
                        chain_row = chain_row[:: int(len(chain_row) / max_points)]
                    log = p in base.log_scales
                    self._plot_walk(
                        ax,
                        p,
                        chain_row,
                        extents=base.extents.get(p),
                        convolve=convolve,
                        color=colors.format(chain.color),
                        log_scale=log,
                    )
            for truth in self.parent._truths:
                if p in truth.location:
                    self._plot_walk_truth(ax, truth, p)

            if p in base.blind:
                ax.set_yticks([])
        else:  # noqa: PLR5501
            if i == 0 and plot_posterior:
                for chain in base.chains:
                    if chain.log_posterior is not None:
                        posterior = chain.log_posterior - chain.log_posterior.max()
                        if len(posterior) > max_points:
                            posterior = posterior[:: int(len(posterior) / max_points)]

                        self._plot_walk(
                            ax,
                            r"$\log(P)$",
                            posterior,
                            convolve=convolve,
                            color=colors.format(chain.color),
                        )
            else:
                label = r"$\log_{10}$Weight" if log_weight else "Weight"

                for chain in base.chains:
                    if chain.weights is not None:
                        weights = chain.weights
                        if len(weights) > max_points:
                            weights = weights[:: int(len(weights) / max_points)]
                        self._plot_walk(
                            ax,
                            label,
                            np.log10(weights) if log_weight else weights,  # type: ignore
                            convolve=convolve,
                            color=colors.format(chain.color),
                        )

    add_watermark(fig, None, fig_size, self.config, size_scale=0.8)
    self._save_fig(fig, filename, dpi=self.config.dpi)

    return fig

plot_distributions

plot_distributions(chains: list[ChainName | Chain] | None = None, columns: list[ColumnName] | None = None, filename: list[str | Path] | str | Path | None = None, col_wrap: int = 4, figsize: float | tuple[float, float] | None = None) -> Figure

Plots the 1D parameter distributions for verification purposes.

This plot is more for a sanity or consistency check than for use with final results. Plotting this before plotting with :func:plot allows you to quickly see if the chains give well behaved distributions, or if certain parameters are suspect or require a greater burn in period.

Parameters:

Name	Type	Description	Default
chains	list[ChainName | Chain] | None	Used to specify which chain to show if more than one chain is loaded in. Can be an integer, specifying the chain index, or a str, specifying the chain name.	None
columns	list[ColumnName] | None	If set, only creates a plot for those specific parameters (if list). If an integer is given, only plots the fist so many parameters.	None
filename	list[str | Path] | str | Path | None	If set, saves the figure to this location	None
figsize	float | tuple[float, float] | None	Scale horizontal and vertical figure size.	None
col_wrap	int	How many columns to plot before wrapping.	4

Returns:

Type	Description
Figure	the matplotlib figure created

Source code in src/chainconsumer/plotter.py

def plot_distributions(
    self,
    chains: list[ChainName | Chain] | None = None,
    columns: list[ColumnName] | None = None,
    filename: list[str | Path] | str | Path | None = None,
    col_wrap: int = 4,
    figsize: float | tuple[float, float] | None = None,
) -> Figure:  # pragma: no cover
    """Plots the 1D parameter distributions for verification purposes.

    This plot is more for a sanity or consistency check than for use with final results.
    Plotting this before plotting with :func:`plot` allows you to quickly see if the
    chains give well behaved distributions, or if certain parameters are suspect
    or require a greater burn in period.

    Args:
        chains:
            Used to specify which chain to show if more than one chain is loaded in.
            Can be an integer, specifying the
            chain index, or a str, specifying the chain name.
        columns:
            If set, only creates a plot for those specific parameters (if list). If an
            integer is given, only plots the fist so many parameters.
        filename:
            If set, saves the figure to this location
        figsize:
            Scale horizontal and vertical figure size.
        col_wrap:
            How many columns to plot before wrapping.

    Returns:
        the matplotlib figure created

    """
    base = self._sanitise(
        chains,
        columns,
        self.config.extents,
        blind=self.config.blind,
        log_scales=self.config.log_scales,
    )

    n = len(base.columns)
    num_cols = min(n, col_wrap)
    num_rows = int(np.ceil(1.0 * n / col_wrap))

    if figsize is None:
        figsize = 1.0
    if isinstance(figsize, float | int):
        figsize_float = figsize
        figsize = (num_cols * 2.5 * figsize, num_rows * 2.5 * figsize)
    else:
        figsize_float = 1.0

    summary = self.config.summarise and len(base.chains) == 1
    hspace = (0.8 if summary else 0.5) / figsize_float
    fig, axes = plt.subplots(nrows=num_rows, ncols=num_cols, figsize=figsize, squeeze=False)
    fig.subplots_adjust(left=0.1, right=0.95, top=0.95, bottom=0.1, wspace=0.05, hspace=hspace)

    formatter = ScalarFormatter(useOffset=False)
    formatter.set_powerlimits((-3, 4))

    for i, ax in enumerate(axes.flatten()):
        if i >= n:
            ax.set_axis_off()
            continue
        p = base.columns[i]

        ax.set_yticks([])
        if p in base.log_scales:
            ax.set_xscale("log")
        if p in base.blind:
            ax.set_xticks([])
        else:
            if self.config.diagonal_tick_labels:
                _ = [label.set_rotation(45) for label in ax.get_xticklabels()]
            _ = [label.set_fontsize(self.config.tick_font_size) for label in ax.get_xticklabels()]

            if p in base.log_scales:
                ax.xaxis.set_major_locator(LogLocator(numticks=self.config.max_ticks))
            else:
                ax.xaxis.set_major_locator(MaxNLocator(self.config.max_ticks, prune="lower"))
                ax.xaxis.set_major_formatter(formatter)
        ax.set_xlim(base.extents.get(p) or self._get_parameter_extents(p, base.chains))

        max_val = -np.inf
        for chain in base.chains:
            if not chain.plot_contour:
                continue
            if p in chain.plotting_columns:
                param_summary = summary and p not in base.blind
                m = self._plot_bars(ax, p, chain, summary=param_summary)
                if max_val is None or m > max_val:
                    max_val = m
        plot_truths(ax, self.parent._truths, py=p)
        ax.set_ylim(0, 1.1 * max_val)
        ax.set_xlabel(p, fontsize=self.config.label_font_size)

    add_watermark(fig, None, figsize, self.config, size_scale=0.8)
    self._save_fig(fig, filename, dpi=self.config.dpi)
    return fig

plot_summary

plot_summary(chains: list[ChainName | Chain] | None = None, columns: list[ColumnName] | None = None, filename: list[str | Path] | str | Path | None = None, figsize: float = 1.0, errorbar: bool = False, extra_parameter_spacing: float = 1.0, vertical_spacing_ratio: float = 1.0) -> Figure

Plots parameter summaries

This plot is more for a sanity or consistency check than for use with final results. Plotting this before plotting with :func:plot allows you to quickly see if the chains give well behaved distributions, or if certain parameters are suspect or require a greater burn in period.

Parameters:

Name	Type	Description	Default
chains	list[ChainName | Chain] | None	Used to specify which chain to show if more than one chain is loaded in. Can be an integer, specifying the chain index, or a str, specifying the chain name.	None
columns	list[ColumnName] | None	If set, only creates a plot for those specific parameters (if list). If an integer is given, only plots the fist so many parameters.	None
filename	list[str | Path] | str | Path | None	If set, saves the figure to this location	None
figsize	float	Scale horizontal and vertical figure size.	1.0
errorbar	bool	Whether to onle plot an error bar, instead of the marginalised distribution.	False
include_truth_chain		If you specify another chain as the truth chain, determine if it should still be plotted.	required
extra_parameter_spacing	float	Increase horizontal space for parameter values	1.0
vertical_spacing_ratio	float	Increase vertical space for each model	1.0

Returns: the matplotlib figure created

Source code in src/chainconsumer/plotter.py

def plot_summary(
    self,
    chains: list[ChainName | Chain] | None = None,
    columns: list[ColumnName] | None = None,
    filename: list[str | Path] | str | Path | None = None,
    figsize: float = 1.0,
    errorbar: bool = False,
    extra_parameter_spacing: float = 1.0,
    vertical_spacing_ratio: float = 1.0,
) -> Figure:  # pragma: no cover
    """Plots parameter summaries

    This plot is more for a sanity or consistency check than for use with final results.
    Plotting this before plotting with :func:`plot` allows you to quickly see if the
    chains give well behaved distributions, or if certain parameters are suspect
    or require a greater burn in period.

    Args:
        chains:
            Used to specify which chain to show if more than one chain is loaded in.
            Can be an integer, specifying the
            chain index, or a str, specifying the chain name.
        columns:
            If set, only creates a plot for those specific parameters (if list). If an
            integer is given, only plots the fist so many parameters.
        filename:
            If set, saves the figure to this location
        figsize:
            Scale horizontal and vertical figure size.
        errorbar:
            Whether to onle plot an error bar, instead of the marginalised distribution.
        include_truth_chain:
            If you specify another chain as the truth chain, determine if it should still
            be plotted.
        extra_parameter_spacing:
            Increase horizontal space for parameter values
        vertical_spacing_ratio:
            Increase vertical space for each model
    Returns:
        the matplotlib figure created

    """
    wide_extents = not errorbar
    base = self._sanitise(
        chains,
        columns,
        self.config.extents,
        blind=self.config.blind,
        log_scales=self.config.log_scales,
        wide_extents=wide_extents,
    )

    # We have a bit of fun to go from chain names to the width of the
    # subplot used to display said names
    max_param = self._get_size_of_texts(base.columns)
    fid_dpi = 65  # Seriously I have no idea what value this should be
    param_width = extra_parameter_spacing + max(0.5, max_param / fid_dpi)
    max_model_name = self._get_size_of_texts([chain.name for chain in base.chains])
    model_width = 0.25 + (max_model_name / fid_dpi)
    gridspec_kw = {
        "width_ratios": [model_width] + [param_width] * len(base.columns),
        "height_ratios": [1] * len(base.chains),
    }
    ncols = 1 + len(base.columns)
    top_spacing = 0.3
    bottom_spacing = 0.2
    row_height = (0.5 if errorbar else 0.8) * vertical_spacing_ratio
    width = param_width * len(base.columns) + model_width
    height = top_spacing + bottom_spacing + row_height * len(base.chains)
    top_ratio = 1 - (top_spacing / height)
    bottom_ratio = bottom_spacing / height

    fig_size = (width * figsize, height * figsize)
    fig, axes = plt.subplots(
        nrows=len(base.chains), ncols=ncols, figsize=fig_size, squeeze=False, gridspec_kw=gridspec_kw
    )
    fig.subplots_adjust(left=0.05, right=0.95, top=top_ratio, bottom=bottom_ratio, wspace=0.0, hspace=0.0)
    label_font_size = self.config.label_font_size
    legend_color_text = self.config.legend_color_text

    max_vals: dict[ColumnName, float] = {}
    num_chains = len(base.chains)
    for i, axes_row in enumerate(axes):
        chain = base.chains[i]
        colour = colors.format(chain.color)

        # First one put name of model
        ax_first = axes_row[0]
        ax_first.set_axis_off()
        text_colour = "k" if not legend_color_text else colour
        ax_first.text(
            0,
            0.5,
            chain.name,
            transform=ax_first.transAxes,
            fontsize=label_font_size,
            verticalalignment="center",
            color=text_colour,
            weight="medium",
        )
        axes_for_summaries = axes_row[1:]

        for ax, p in zip(axes_for_summaries, base.columns):
            # Set up the frames
            if i > 0:
                ax.spines["top"].set_visible(False)
            if i < (num_chains - 1):
                ax.spines["bottom"].set_visible(False)
            if i < (num_chains - 1) or p in base.blind:
                ax.set_xticks([])
            ax.set_yticks([])
            ax.set_xlim(base.extents[p])
            if p in base.log_scales:
                ax.set_xscale("log")

            # Put title in
            if i == 0:
                ax.set_title(self.config.get_label(p), fontsize=label_font_size)

            # Add truth values
            for truth in self.parent._truths:
                truth_value = truth.location.get(p)
                if truth_value is not None:
                    ax.axvline(truth_value, **truth._kwargs)

            # Skip if this chain doesnt have the parameter
            if p not in chain.data_columns:
                continue

            # Plot the good stuff
            if errorbar:
                fv = self.parent.analysis.get_parameter_summary(chain, p)
                if fv is None or fv.all_none:
                    continue
                if fv.lower is not None and fv.upper is not None:
                    diff = np.abs(np.diff(fv.array))
                    ax.errorbar([fv.center], 0, xerr=[[diff[0]], [diff[1]]], fmt="o", color=colour)
            else:
                m = self._plot_bars(ax, p, chain)
                if max_vals.get(p) is None or m > max_vals[p]:
                    max_vals[p] = m

    for i, axes_row in enumerate(axes):
        for ax, p in zip(axes_row, base.columns):
            if not errorbar:
                ax.set_ylim(0, 1.1 * max_vals[p])

    add_watermark(fig, None, fig_size, self.config, size_scale=0.8)
    self._save_fig(fig, filename, dpi=self.config.dpi)

    return fig