Source code for scqubits.utils.plotting

# plotting.py
#
# This file is part of scqubits.
#
#    Copyright (c) 2019 and later, Jens Koch and Peter Groszkowski
#    All rights reserved.
#
#    This source code is licensed under the BSD-style license found in the
#    LICENSE file in the root directory of this source tree.
############################################################################

import functools
import operator
import os
import warnings

from typing import (
    TYPE_CHECKING,
    Any,
    Callable,
    Dict,
    Iterable,
    List,
    Optional,
    Tuple,
    Union,
)

import matplotlib as mpl
import matplotlib.pyplot as plt
import numpy as np

from matplotlib.axes import Axes
from matplotlib.figure import Figure
from mpl_toolkits.axes_grid1 import make_axes_locatable

import scqubits.core.constants as constants
import scqubits.settings as settings
import scqubits.utils.misc as utils
import scqubits.utils.plot_defaults as defaults

if TYPE_CHECKING:
    from scqubits.core.storage import SpectrumData, WaveFunction, WaveFunctionOnGrid

try:
    from labellines import labelLines

    _LABELLINES_ENABLED = True
except ImportError:
    _LABELLINES_ENABLED = False


# A dictionary of plotting options that are directly passed to specific matplotlib's
# plot commands.
_direct_plot_options = {
    "plot": ("alpha", "color", "linestyle", "linewidth", "marker", "markersize"),
    "imshow": ("interpolation",),
    "contourf": tuple(),  # empty for now
}


def _extract_kwargs_options(
    kwargs: Dict[str, Any], plot_type: str, direct_plot_options: Dict[str, Any] = None
) -> Dict[str, Any]:
    """
    Select options from kwargs for a given plot_type and return them in a dictionary.

    Parameters
    ----------
    kwargs:
        dictionary with options that can be passed to different plotting commands
    plot_type:
        a type of plot for which the options should be selected
    direct_plot_options:
        a lookup dictionary with supported options for a given plot_type

    Returns
    ----------
        dictionary with key/value pairs corresponding to selected options from kwargs

    """
    direct_plot_options = direct_plot_options or _direct_plot_options
    d = {}
    if plot_type in direct_plot_options:
        for key in kwargs:
            if key in direct_plot_options[plot_type]:
                d[key] = kwargs[key]
    return d


def _process_options(
    figure: Figure, axes: Axes, opts: Dict[str, Any] = None, **kwargs
) -> None:
    """
    Processes plotting options.

    Parameters
    ----------
    figure:
    axes:
    opts:
        keyword dictionary with custom options
    **kwargs:
        standard plotting option (see separate documentation)
    """
    opts = opts or {}

    # Only process items in kwargs that would not have been
    # processed through _extract_kwargs_options()
    filtered_kwargs = {
        key: value
        for key, value in kwargs.items()
        if key not in functools.reduce(operator.concat, _direct_plot_options.values())
    }  # type: ignore

    option_dict = {**opts, **filtered_kwargs}

    for key, value in option_dict.items():
        if key in defaults.SPECIAL_PLOT_OPTIONS:
            _process_special_option(figure, axes, key, value)
        else:
            set_method = getattr(axes, "set_" + key)
            set_method(value)

    filename = kwargs.get("filename")
    if filename:
        figure.savefig(os.path.splitext(filename)[0] + ".pdf")

    if settings.DESPINE and not axes.name == "3d":
        # Hide the right and top spines
        axes.spines["right"].set_visible(False)
        axes.spines["top"].set_visible(False)

        # Only show ticks on the left and bottom spines
        axes.yaxis.set_ticks_position("left")
        axes.xaxis.set_ticks_position("bottom")


def _process_special_option(figure: Figure, axes: Axes, key: str, value: Any) -> None:
    """Processes a single 'special' option, i.e., one internal to scqubits and not to be handed further down to
    matplotlib.
    """
    if key == "ymax":
        ymax = value
        ymin, _ = axes.get_ylim()
        ymin = ymin - (ymax - ymin) * 0.05
        axes.set_ylim(ymin, ymax)
    elif key == "figsize":
        figure.set_size_inches(value)
    elif key == "grid":
        axes.grid(**value) if isinstance(value, dict) else axes.grid(value)


[docs]def wavefunction1d( wavefuncs: Union["WaveFunction", "List[WaveFunction]"], potential_vals: np.ndarray = None, offset: Union[float, Iterable[float]] = 0, scaling: Optional[float] = None, **kwargs ) -> Tuple[Figure, Axes]: """ Plots the amplitude of a single real-valued 1d wave function, along with the potential energy if provided. Parameters ---------- wavefuncs: basis and amplitude data of wave function to be plotted potential_vals: potential energies, array length must match basis array of `wavefunc` offset: y-offset for the wave function (e.g., shift by eigenenergy) scaling: scaling factor for wave function amplitudes **kwargs: standard plotting option (see separate documentation) Returns ------- matplotlib objects for further editing """ fig, axes = kwargs.get("fig_ax") or plt.subplots() offset_list = [offset] if not isinstance(offset, (list, np.ndarray)) else offset wavefunc_list = [wavefuncs] if not isinstance(wavefuncs, list) else wavefuncs scale_constant = renormalization_factor(wavefunc_list[0], potential_vals) for wavefunc in wavefunc_list: wavefunc.amplitudes *= scale_constant scale_factor = scaling or defaults.set_wavefunction_scaling( wavefunc_list, potential_vals ) for wavefunction, energy_offset in zip(wavefunc_list, offset_list): x_vals = wavefunction.basis_labels y_vals = energy_offset + scale_factor * wavefunction.amplitudes offset_vals = [energy_offset] * len(x_vals) axes.plot(x_vals, y_vals, **_extract_kwargs_options(kwargs, "plot")) axes.fill_between( x_vals, y_vals, offset_vals, where=(y_vals != offset_vals), interpolate=True ) if potential_vals is not None: y_min = np.min(potential_vals) y_max = np.max(offset_list) y_range = y_max - y_min y_max += 0.3 * y_range y_min = np.min(potential_vals) - 0.1 * y_range axes.set_ylim([y_min, y_max]) axes.plot( x_vals, potential_vals, color="gray", **_extract_kwargs_options(kwargs, "plot") ) _process_options(fig, axes, **kwargs) return fig, axes
[docs]def renormalization_factor( wavefunc: "WaveFunction", potential_vals: np.ndarray ) -> float: """ Takes the amplitudes of one wavefunction and the potential values to scale the dimensionless amplitude to a (pseudo-)energy that allows us to plot wavefunctions and energies in the same plot. Parameters ---------- wavefunc: ndarray of wavefunction amplitudes potential_vals: array of potential energy values (that determine the energy range on the y axis Returns ------- renormalization factor that converts the wavefunction amplitudes into energy units """ FILL_FACTOR = 0.1 energy_range = np.max(potential_vals) - np.min(potential_vals) amplitude_range = np.max(wavefunc.amplitudes) - np.min(wavefunc.amplitudes) if amplitude_range < 1.0e-10: return 0.0 return FILL_FACTOR * energy_range / amplitude_range
[docs]def wavefunction1d_discrete(wavefunc: "WaveFunction", **kwargs) -> Tuple[Figure, Axes]: """ Plots the amplitude of a real-valued 1d wave function in a discrete basis. (Example: transmon in the charge basis.) Parameters ---------- wavefunc: basis and amplitude data of wave function to be plotted **kwargs: standard plotting option (see separate documentation) Returns ------- matplotlib objects for further editing """ fig, axes = kwargs.get("fig_ax") or plt.subplots() x_vals = wavefunc.basis_labels width = 0.75 axes.bar(x_vals, wavefunc.amplitudes, width=width) axes.set_xticks(x_vals) axes.set_xticklabels(x_vals) _process_options(fig, axes, defaults.wavefunction1d_discrete(), **kwargs) return fig, axes
[docs]def wavefunction2d( wavefunc: "WaveFunctionOnGrid", zero_calibrate: bool = False, **kwargs ) -> Tuple[Figure, Axes]: """ Creates a density plot of the amplitude of a real-valued wave function in 2 "spatial" dimensions. Parameters ---------- wavefunc: basis and amplitude data of wave function to be plotted zero_calibrate: whether to calibrate plot to zero amplitude **kwargs: standard plotting option (see separate documentation) Returns ------- matplotlib objects for further editing """ fig, axes = kwargs.get("fig_ax") or plt.subplots() min_vals = wavefunc.gridspec.min_vals max_vals = wavefunc.gridspec.max_vals if zero_calibrate: absmax = np.amax(np.abs(wavefunc.amplitudes)) imshow_minval = -absmax imshow_maxval = absmax cmap = plt.get_cmap("PRGn") else: imshow_minval = np.min(wavefunc.amplitudes) imshow_maxval = np.max(wavefunc.amplitudes) cmap = plt.cm.viridis im = axes.imshow( wavefunc.amplitudes, extent=[min_vals[0], max_vals[0], min_vals[1], max_vals[1]], cmap=cmap, vmin=imshow_minval, vmax=imshow_maxval, origin="lower", aspect="auto", **_extract_kwargs_options(kwargs, "imshow") ) divider = make_axes_locatable(axes) cax = divider.append_axes("right", size="2%", pad=0.05) fig.colorbar(im, cax=cax) _process_options(fig, axes, defaults.wavefunction2d(), **kwargs) return fig, axes
[docs]def contours( x_vals: Iterable[float], y_vals: Iterable[float], func: Callable, contour_vals: Iterable[float] = None, show_colorbar: bool = True, **kwargs ) -> Tuple[Figure, Axes]: """Contour plot of a 2d function `func(x,y)`. Parameters ---------- x_vals: x values for the x-y evaluation grid y_vals: y values for the x-y evaluation grid func: function f(x,y) for which contours are to be plotted contour_vals: contour values can be specified if so desired show_colorbar: **kwargs: standard plotting option (see separate documentation) Returns ------- matplotlib objects for further editing """ fig, axes = kwargs.get("fig_ax") or plt.subplots() x_grid, y_grid = np.meshgrid(x_vals, y_vals) z_array = func(x_grid, y_grid) im = axes.contourf( x_grid, y_grid, z_array, levels=contour_vals, cmap=plt.cm.viridis, origin="lower", **_extract_kwargs_options(kwargs, "contourf") ) if show_colorbar: divider = make_axes_locatable(axes) cax = divider.append_axes("right", size="2%", pad=0.05) fig.colorbar(im, cax=cax) _process_options(fig, axes, opts=defaults.contours(x_vals, y_vals), **kwargs) return fig, axes
[docs]def matrix( data_matrix: np.ndarray, mode: str = "abs", show_numbers: bool = False, **kwargs ) -> Tuple[Figure, Tuple[Axes, Axes]]: """ Create a "skyscraper" plot and a 2d color-coded plot of a matrix. Parameters ---------- data_matrix: 2d matrix data mode: choice from `constants.MODE_FUNC_DICT` for processing function to be applied to data show_numbers: determines whether matrix element values are printed on top of the plot (default: False) **kwargs: standard plotting option (see separate documentation) Returns ------- figure and axes objects for further editing """ if "fig_ax" in kwargs: fig, (ax1, ax2) = kwargs["fig_ax"] else: fig = plt.figure() ax1 = fig.add_subplot(1, 2, 1, projection="3d") ax2 = plt.subplot(1, 2, 2) fig, ax2 = matrix2d( data_matrix, mode=mode, show_numbers=show_numbers, fig_ax=(fig, ax2), **kwargs ) fig, ax1 = matrix_skyscraper(data_matrix, mode=mode, fig_ax=(fig, ax1), **kwargs) return fig, (ax1, ax2)
[docs]def matrix_skyscraper( matrix: np.ndarray, mode: str = "abs", **kwargs ) -> Tuple[Figure, Axes]: """Display a 3d skyscraper plot of the matrix Parameters ---------- matrix: 2d matrix data mode: choice from `constants.MODE_FUNC_DICT` for processing function to be applied to data **kwargs: standard plotting option (see separate documentation) Returns ------- figure and axes objects for further editing """ fig, axes = kwargs.get("fig_ax") or plt.subplots(projection="3d") y_count, x_count = matrix.shape # We label the columns as "x", while rows as "y" element_count = x_count * y_count # total num. of elements to plot xgrid, ygrid = np.meshgrid(range(x_count), range(y_count)) xgrid = xgrid.flatten() ygrid = ygrid.flatten() zbottom = np.zeros(element_count) # all bars start at z=0 dx, dy = 0.75, 0.75 # width of bars in x and y directions modefunction = constants.MODE_FUNC_DICT[mode] zheight = modefunction(matrix).flatten() # height of bars from matrix elements min_zheight, max_zheight = min(zheight), max(zheight) if mode == "abs" or mode == "abs_sqr": nrm = mpl.colors.Normalize( 0, max_zheight ) # normalize colors between 0 and max. data else: nrm = mpl.colors.Normalize( min_zheight, max_zheight ) # normalize colors between min. and max. of data colors = plt.cm.viridis(nrm(zheight)) # list of colors for each bar # skyscraper plot axes.view_init(azim=210, elev=23) axes.bar3d(xgrid, ygrid, zbottom, dx, dy, zheight, color=colors) if mode == "abs" or mode == "abs_sqr": min_z, max_z = 0, max_zheight else: # mode is "real" or "imag" min_z = 0 if min_zheight > 0 else min_zheight max_z = 0 if max_zheight < 0 else max_zheight if min_z == max_z: # pad with small values so we don't get warnings max_z += 0.0000001 axes.set_zlim3d([min_z, max_z]) for axis, locs in [ (axes.xaxis, np.arange(x_count)), (axes.yaxis, np.arange(y_count)), ]: axis.set_ticks(locs + 0.5, minor=True) axis.set(ticks=locs + 0.5, ticklabels=locs) _process_options(fig, axes, opts=defaults.matrix(), **kwargs) return fig, axes
[docs]def matrix2d( matrix: np.ndarray, mode: str = "abs", show_numbers: bool = True, **kwargs ) -> Tuple[Figure, Axes]: """Display a matrix as a color-coded 2d plot, optionally printing the numerical values of the matrix elements. Parameters ---------- matrix: 2d matrix data mode: choice from `constants.MODE_FUNC_DICT` for processing function to be applied to data show_numbers: determines whether matrix element values are printed on top of the plot (default: True) **kwargs: standard plotting option (see separate documentation) Returns ------- figure and axes objects for further editing """ fig, axes = kwargs.get("fig_ax") or plt.subplots() modefunction = constants.MODE_FUNC_DICT[mode] zheight = modefunction(matrix).flatten() # height of bars from matrix elements if mode == "abs" or mode == "abs_sqr": nrm = mpl.colors.Normalize( 0, max(zheight) ) # normalize colors between 0 and max. data else: nrm = mpl.colors.Normalize( min(zheight), max(zheight) ) # normalize colors between min. and max. of data axes.matshow(modefunction(matrix), cmap=plt.cm.viridis, interpolation=None) cax, _ = mpl.colorbar.make_axes( axes, shrink=0.75, pad=0.02 ) # add colorbar with normalized range mpl.colorbar.ColorbarBase(cax, cmap=plt.cm.viridis, norm=nrm) if show_numbers: for y_index in range(matrix.shape[0]): for x_index in range(matrix.shape[1]): axes.text( x_index, y_index, "{:.03f}".format(modefunction(matrix[y_index, x_index])), va="center", ha="center", fontsize=8, rotation=45, color="white", ) # shift the grid for axis, locs in [ (axes.xaxis, np.arange(matrix.shape[1])), (axes.yaxis, np.arange(matrix.shape[0])), ]: axis.set_ticks(locs + 0.5, minor=True) axis.set(ticks=locs, ticklabels=locs) axes.grid(False) _process_options(fig, axes, **kwargs) axes.tick_params(axis="x", bottom=False, top=True, labelbottom=False, labeltop=True) return fig, axes
[docs]def data_vs_paramvals( xdata: np.ndarray, ydata: np.ndarray, label_list: Union[List[str], List[int]] = None, **kwargs ) -> Tuple[Figure, Axes]: """Plot of a set of yadata vs xdata. The individual points correspond to the a provided array of parameter values. Parameters ---------- xdata, ydata: must have compatible shapes for matplotlib.pyplot.plot label_list: list of labels associated with the individual curves to be plotted **kwargs: standard plotting option (see separate documentation) Returns ------- matplotlib objects for further editing """ fig, axes = kwargs.get("fig_ax") or plt.subplots() if label_list is None: axes.plot(xdata, ydata, **_extract_kwargs_options(kwargs, "plot")) else: for idx, ydataset in enumerate(ydata.T): axes.plot( xdata, ydataset, label=label_list[idx], **_extract_kwargs_options(kwargs, "plot") ) if _LABELLINES_ENABLED: try: labelLines(axes.get_lines(), zorder=2.0) except Exception: pass else: axes.legend(loc="center left", bbox_to_anchor=(1, 0.5)) _process_options(fig, axes, **kwargs) return fig, axes
[docs]def evals_vs_paramvals( specdata: "SpectrumData", which: Union[int, Iterable[int]] = -1, subtract_ground: bool = False, label_list: List[str] = None, **kwargs ) -> Tuple[Figure, Axes]: """Generates a simple plot of a set of eigenvalues as a function of one parameter. The individual points correspond to the a provided array of parameter values. Parameters ---------- specdata: object includes parameter name, values, and resulting eigenenergies which: number of desired eigenvalues (sorted from smallest to largest); default: -1, signals all eigenvalues or: list of specific eigenvalues to include subtract_ground: whether to subtract the ground state energy label_list: list of labels associated with the individual curves to be plotted **kwargs: standard plotting option (see separate documentation) Returns ------- matplotlib objects for further editing """ index_list = utils.process_which(which, specdata.energy_table[0].size) xdata = specdata.param_vals ydata = specdata.energy_table[:, index_list] if subtract_ground: ydata = (ydata.T - ydata[:, 0]).T return data_vs_paramvals( xdata, ydata, label_list=label_list, **defaults.evals_vs_paramvals(specdata, **kwargs) )
[docs]def matelem_vs_paramvals( specdata: "SpectrumData", select_elems: Union[int, List[Tuple[int, int]]] = 4, mode: str = "abs", **kwargs ) -> Tuple[Figure, Axes]: """Generates a simple plot of matrix elements as a function of one parameter. The individual points correspond to the a provided array of parameter values. Parameters ---------- specdata: object includes parameter name, values, and matrix elements select_elems: either maximum index of desired matrix elements, or list [(i1, i2), (i3, i4), ...] of index tuples for specific desired matrix elements mode: choice of processing function to be applied to data (default value = 'abs') **kwargs: standard plotting option (see separate documentation) Returns ------- matplotlib objects for further editing """ fig, axes = kwargs.get("fig_ax") or plt.subplots() x = specdata.param_vals modefunction = constants.MODE_FUNC_DICT[mode] if isinstance(select_elems, int): index_pairs = [ (row, col) for row in range(select_elems) for col in range(row + 1) ] else: index_pairs = select_elems for (row, col) in index_pairs: y = modefunction(specdata.matrixelem_table[:, row, col]) axes.plot( x, y, label=str(row) + "," + str(col), **_extract_kwargs_options(kwargs, "plot") ) if _LABELLINES_ENABLED: labelLines(axes.get_lines(), zorder=1.5) else: axes.legend(loc="center left", bbox_to_anchor=(1, 0.5)) _process_options(fig, axes, opts=defaults.matelem_vs_paramvals(specdata), **kwargs) return fig, axes