visualization.py

# -*- coding: utf-8 -*-
from collections import namedtuple
import warnings

import numpy as np

import meep as mp
from meep.geom import Vector3, init_do_averaging
from meep.source import EigenModeSource, check_positive


# ------------------------------------------------------- #
# Visualization
# ------------------------------------------------------- #
# Contains all necesarry visualation routines for use with
# pymeep and pympb.

# ------------------------------------------------------- #
# Functions used to define the default plotting parameters
# for the different plotting routines.

default_source_parameters = {
        'color':'r',
        'edgecolor':'r',
        'facecolor':'none',
        'hatch':'/',
        'linewidth':2
    }

default_monitor_parameters = {
        'color':'b',
        'edgecolor':'b',
        'facecolor':'none',
        'hatch':'/',
        'linewidth':2
    }

default_field_parameters = {
        'interpolation':'spline36',
        'cmap':'RdBu',
        'alpha':0.6,
        'post_process':np.real
        }

default_eps_parameters = {
        'interpolation':'spline36',
        'cmap':'binary',
        'alpha':1.0,
        'contour':False,
        'contour_linewidth':1
    }

default_boundary_parameters = {
        'color':'g',
        'edgecolor':'g',
        'facecolor':'none',
        'hatch':'/'
    }

default_volume_parameters = {
        'alpha':1.0,
        'color':'k',
        'linestyle':'-',
        'linewidth':1,
        'marker':'.',
        'edgecolor':'k',
        'facecolor':'none',
        'hatch':'/'
    }

default_label_parameters = {
    'label_color':'r',
    'offset':20,
    'label_alpha':0.3
}

# Used to remove the elements of a dictionary (dict_to_filter) that
# don't correspond to the keyword arguments of a particular
# function (func_with_kwargs.)
# Adapted from https://stackoverflow.com/questions/26515595/how-does-one-ignore-unexpected-keyword-arguments-passed-to-a-function/44052550
def filter_dict(dict_to_filter, func_with_kwargs):
    import inspect
    filter_keys = []
    try:
        # Python3 ...
        sig = inspect.signature(func_with_kwargs)
        filter_keys = [param.name for param in sig.parameters.values() if param.kind == param.POSITIONAL_OR_KEYWORD]
    except:
        # Python2 ...
        filter_keys = inspect.getargspec(func_with_kwargs)[0]

    filtered_dict = {filter_key:dict_to_filter[filter_key] for filter_key in filter_keys if filter_key in dict_to_filter}
    return filtered_dict

# ------------------------------------------------------- #
# Routines to add legends to plot

def place_label(ax,label_text,x,y,centerx,centery,label_parameters=None):

    label_parameters = default_label_parameters if label_parameters is None else dict(default_label_parameters, **label_parameters)

    offset = label_parameters['offset']
    alpha = label_parameters['label_alpha']
    color = label_parameters['label_color']

    if x > centerx:
        xtext = -offset
    else:
        xtext = offset
    if y > centery:
        ytext = -offset
    else:
        ytext = offset

    ax.annotate(label_text, xy=(x, y), xytext=(xtext,ytext),
            textcoords='offset points', ha='center', va='bottom',
            bbox=dict(boxstyle='round,pad=0.2', fc=color, alpha=alpha),
            arrowprops=dict(arrowstyle='->', connectionstyle='arc3,rad=0.5',
                            color=color))
    return ax

# ------------------------------------------------------- #
# Helper functions used to plot volumes on a 2D plane

# Returns the intersection points of 2 Volumes.
# Volumes must be a line, plane, or rectangular prism
# (since they are volume objects)
def intersect_volume_volume(volume1,volume2):
    # volume1 ............... [volume]
    # volume2 ............... [volume]

    # Represent the volumes by an "upper" and "lower" coordinate
    U1 = [volume1.center.x+volume1.size.x/2,volume1.center.y+volume1.size.y/2,volume1.center.z+volume1.size.z/2]
    L1 = [volume1.center.x-volume1.size.x/2,volume1.center.y-volume1.size.y/2,volume1.center.z-volume1.size.z/2]

    U2 = [volume2.center.x+volume2.size.x/2,volume2.center.y+volume2.size.y/2,volume2.center.z+volume2.size.z/2]
    L2 = [volume2.center.x-volume2.size.x/2,volume2.center.y-volume2.size.y/2,volume2.center.z-volume2.size.z/2]

    # Evaluate intersection
    U = np.min([U1,U2],axis=0)
    L = np.max([L1,L2],axis=0)

    # For single points we have to check manually
    if np.all(U-L == 0):
        if (not volume1.pt_in_volume(Vector3(*U))) or (not volume2.pt_in_volume(Vector3(*U))):
            return []

    # Check for two volumes that don't intersect
    if np.any(U-L < 0):
        return []

    # Pull all possible vertices
    vertices = []
    for x_vals in [L[0],U[0]]:
        for y_vals in [L[1],U[1]]:
            for z_vals in [L[2],U[2]]:
                vertices.append(Vector3(x_vals,y_vals,z_vals))

    # Remove any duplicate points caused by coplanar lines
    vertices = [vertices[i] for i, x in enumerate(vertices) if x not in vertices[i+1:]]

    return vertices

# All of the 2D plotting routines need an output plane over which to plot.
# The user has many options to specify this output plane. They can pass
# the output_plane parameter, which is a 2D volume object. They can specify
# a volume using in_volume, which stores the volume as a C volume, not a python
# volume. They can also do nothing and plot the XY plane through Z=0.
#
# Not only do we need to check for all of these possibilities, but we also need
# to check if the user accidentally specifies a plane that stretches beyond the
# simulation domain.
def get_2D_dimensions(sim,output_plane):
    from meep.simulation import Volume

    # Pull correct plane from user
    if output_plane:
        plane_center, plane_size = (output_plane.center, output_plane.size)
    elif sim.output_volume:
        plane_center, plane_size = mp.get_center_and_size(sim.output_volume)
    else:
        plane_center, plane_size = (sim.geometry_center, sim.cell_size)
    plane_volume = Volume(center=plane_center,size=plane_size)

    if plane_size.x!=0 and plane_size.y!=0 and plane_size.z!=0:
        raise ValueError("Plane volume must be 2D (a plane).")

    check_volume = Volume(center=sim.geometry_center,size=sim.cell_size)

    vertices = intersect_volume_volume(check_volume,plane_volume)

    if len(vertices) == 0:
        raise ValueError("The specified user volume is completely outside of the simulation domain.")

    intersection_vol = Volume(vertices=vertices)

    if (intersection_vol.size != plane_volume.size) or (intersection_vol.center != plane_volume.center):
        warnings.warn('The specified user volume is larger than the simulation domain and has been truncated.')

    sim_center, sim_size = (intersection_vol.center, intersection_vol.size)
    return sim_center, sim_size

# ------------------------------------------------------- #
# actual plotting routines

def plot_volume(sim,ax,volume,output_plane=None,plotting_parameters=None,label=None):
    if not sim._is_initialized:
        sim.init_sim()

    import matplotlib.patches as patches
    from matplotlib import pyplot as plt
    from meep.simulation import Volume

    # Set up the plotting parameters
    plotting_parameters = default_volume_parameters if plotting_parameters is None else dict(default_volume_parameters, **plotting_parameters)

    # Get domain measurements
    sim_center, sim_size = get_2D_dimensions(sim,output_plane)

    plane = Volume(center=sim_center,size=sim_size)

    # Pull volume parameters
    size = volume.size
    center = volume.center

    xmax = center.x+size.x/2
    xmin = center.x-size.x/2
    ymax = center.y+size.y/2
    ymin = center.y-size.y/2
    zmax = center.z+size.z/2
    zmin = center.z-size.z/2

    # Add labels if requested
    if label is not None and mp.am_master():
        if sim_size.x == 0:
            ax = place_label(ax,label,center.y,center.z,sim_center.y,sim_center.z,label_parameters=plotting_parameters)
        elif sim_size.y == 0:
            ax = place_label(ax,label,center.x,center.z,sim_center.x,sim_center.z,label_parameters=plotting_parameters)
        elif sim_size.z == 0:
            ax = place_label(ax,label,center.x,center.y,sim_center.x,sim_center.y,label_parameters=plotting_parameters)

    # Intersect plane with volume
    intersection = intersect_volume_volume(volume,plane)

    # Sort the points in a counter clockwise manner to ensure convex polygon is formed
    def sort_points(xy):
        xy = np.squeeze(xy)
        xy_mean = np.mean(xy,axis=0)
        theta = np.arctan2(xy[:,1]-xy_mean[1],xy[:,0]-xy_mean[0])
        return xy[np.argsort(theta,axis=0),:]

    if mp.am_master():
        # Point volume
        if len(intersection) == 1:
            point_args = {key:value for key, value in plotting_parameters.items() if key in ['color','marker','alpha','linewidth']}
            if sim_size.y==0:
                ax.scatter(center.x,center.z, **point_args)
                return ax
            elif sim_size.x==0:
                ax.scatter(center.y,center.z, **point_args)
                return ax
            elif sim_size.z==0:
                ax.scatter(center.x,center.y, **point_args)
                return ax
            else:
                return ax

        # Line volume
        elif len(intersection) == 2:
            line_args = {key:value for key, value in plotting_parameters.items() if key in ['color','linestyle','linewidth','alpha']}
            # Plot YZ
            if sim_size.x==0:
                ax.plot([a.y for a in intersection],[a.z for a in intersection], **line_args)
                return ax
            #Plot XZ
            elif sim_size.y==0:
                ax.plot([a.x for a in intersection],[a.z for a in intersection], **line_args)
                return ax
            # Plot XY
            elif sim_size.z==0:
                ax.plot([a.x for a in intersection],[a.y for a in intersection], **line_args)
                return ax
            else:
                return ax

        # Planar volume
        elif len(intersection) > 2:
            planar_args = {key:value for key, value in plotting_parameters.items() if key in ['edgecolor','linewidth','facecolor','hatch','alpha']}
            # Plot YZ
            if sim_size.x==0:
                ax.add_patch(patches.Polygon(sort_points([[a.y,a.z] for a in intersection]), **planar_args))
                return ax
            #Plot XZ
            elif sim_size.y==0:
                ax.add_patch(patches.Polygon(sort_points([[a.x,a.z] for a in intersection]), **planar_args))
                return ax
            # Plot XY
            elif sim_size.z==0:
                ax.add_patch(patches.Polygon(sort_points([[a.x,a.y] for a in intersection]), **planar_args))
                return ax
            else:
                return ax
        else:
            return ax
    return ax

def plot_eps(sim,ax,output_plane=None,eps_parameters=None,frequency=0):
    if sim.structure is None:
        sim.init_sim()


    # consolidate plotting parameters
    eps_parameters = default_eps_parameters if eps_parameters is None else dict(default_eps_parameters, **eps_parameters)

    # Get domain measurements
    sim_center, sim_size = get_2D_dimensions(sim,output_plane)

    xmin = sim_center.x - sim_size.x/2
    xmax = sim_center.x + sim_size.x/2
    ymin = sim_center.y - sim_size.y/2
    ymax = sim_center.y + sim_size.y/2
    zmin = sim_center.z - sim_size.z/2
    zmax = sim_center.z + sim_size.z/2

    center = Vector3(sim_center.x,sim_center.y,sim_center.z)
    cell_size = Vector3(sim_size.x,sim_size.y,sim_size.z)

    if sim_size.x == 0:
        # Plot y on x axis, z on y axis (YZ plane)
        extent = [ymin,ymax,zmin,zmax]
        xlabel = 'Y'
        ylabel = 'Z'
    elif sim_size.y == 0:
        # Plot x on x axis, z on y axis (XZ plane)
        extent = [xmin,xmax,zmin,zmax]
        xlabel = 'X'
        ylabel = 'Z'
    elif sim_size.z == 0:
        # Plot x on x axis, y on y axis (XY plane)
        extent = [xmin,xmax,ymin,ymax]
        xlabel = 'X'
        ylabel = 'Y'
    else:
        raise ValueError("A 2D plane has not been specified...")

    eps_data = np.rot90(np.real(sim.get_array(center=center, size=cell_size, component=mp.Dielectric, frequency=frequency)))
    if mp.am_master():
        if eps_parameters['contour']:
            ax.contour(eps_data, 0, colors='black', origin='upper', extent=extent, linewidths=eps_parameters['contour_linewidth'])
        else:
            ax.imshow(eps_data, extent=extent, **filter_dict(eps_parameters, ax.imshow))
        ax.set_xlabel(xlabel)
        ax.set_ylabel(ylabel)

    return ax

def plot_boundaries(sim,ax,output_plane=None,boundary_parameters=None):
    if not sim._is_initialized:
        sim.init_sim()

    # consolidate plotting parameters
    boundary_parameters = default_boundary_parameters if boundary_parameters is None else dict(default_boundary_parameters, **boundary_parameters)

    def get_boundary_volumes(thickness,direction,side):
        from meep.simulation import Volume

        thickness = boundary.thickness

        # Get domain measurements
        sim_center, sim_size = (sim.geometry_center, sim.cell_size)

        xmin = sim_center.x - sim_size.x/2
        xmax = sim_center.x + sim_size.x/2
        ymin = sim_center.y - sim_size.y/2
        ymax = sim_center.y + sim_size.y/2
        zmin = sim_center.z - sim_size.z/2
        zmax = sim_center.z + sim_size.z/2

        cell_x = sim.cell_size.x
        cell_y = sim.cell_size.y
        cell_z = sim.cell_size.z

        if direction == mp.X and side == mp.Low:
            return Volume(center=Vector3(xmin+thickness/2,sim.geometry_center.y,sim.geometry_center.z),
            size=Vector3(thickness,cell_y,cell_z))
        elif direction == mp.X and side == mp.High:
            return Volume(center=Vector3(xmax-thickness/2,sim.geometry_center.y,sim.geometry_center.z),
            size=Vector3(thickness,cell_y,cell_z))
        elif direction == mp.Y and side == mp.Low:
            return Volume(center=Vector3(sim.geometry_center.x,ymin+thickness/2,sim.geometry_center.z),
            size=Vector3(cell_x,thickness,cell_z))
        elif direction == mp.Y and side == mp.High:
            return Volume(center=Vector3(sim.geometry_center.x,ymax-thickness/2,sim.geometry_center.z),
            size=Vector3(cell_x,thickness,cell_z))
        elif direction == mp.Z and side == mp.Low:
            return Volume(center=Vector3(sim.geometry_center.x,sim.geometry_center.y,zmin+thickness/2),
            size=Vector3(cell_x,cell_y,thickness))
        elif direction == mp.Z and side == mp.High:
            return Volume(center=Vector3(sim.geometry_center.x,sim.geometry_center.y,zmax-thickness/2),
            size=Vector3(cell_x,cell_y,thickness))
        else:
            raise ValueError("Invalid boundary type")

    import itertools
    for boundary in sim.boundary_layers:
        # All 4 side are the same
        if boundary.direction == mp.ALL and boundary.side == mp.ALL:
            if sim.dimensions == 1:
                dims = [mp.X]
            elif sim.dimensions == 2:
                dims = [mp.X,mp.Y]
            elif sim.dimensions == 3:
                dims = [mp.X,mp.Y,mp.Z]
            else:
                raise ValueError("Invalid simulation dimensions")
            for permutation in itertools.product(dims, [mp.Low, mp.High]):
                vol = get_boundary_volumes(boundary.thickness,*permutation)
                ax = plot_volume(sim,ax,vol,output_plane,plotting_parameters=boundary_parameters)
        # 2 sides are the same
        elif boundary.side == mp.ALL:
            for side in [mp.Low, mp.High]:
                vol = get_boundary_volumes(boundary.thickness,boundary.direction,side)
                ax = plot_volume(sim,ax,vol,output_plane,plotting_parameters=boundary_parameters)
        # only one side
        else:
            vol = get_boundary_volumes(boundary.thickness,boundary.direction,boundary.side)
            ax = plot_volume(sim,ax,vol,output_plane,plotting_parameters=boundary_parameters)
    return ax

def plot_sources(sim,ax,output_plane=None,labels=False,source_parameters=None):
    if not sim._is_initialized:
        sim.init_sim()

    from meep.simulation import Volume

    # consolidate plotting parameters
    source_parameters = default_source_parameters if source_parameters is None else dict(default_source_parameters, **source_parameters)

    label = 'source' if labels else None

    for src in sim.sources:
        vol = Volume(center=src.center,size=src.size)
        ax = plot_volume(sim,ax,vol,output_plane,plotting_parameters=source_parameters,label=label)
    return ax

def plot_monitors(sim,ax,output_plane=None,labels=False,monitor_parameters=None):
    if not sim._is_initialized:
        sim.init_sim()

    from meep.simulation import Volume

     # consolidate plotting parameters
    monitor_parameters = default_monitor_parameters if monitor_parameters is None else dict(default_monitor_parameters, **monitor_parameters)

    label = 'monitor' if labels else None

    for mon in sim.dft_objects:
        for reg in mon.regions:
            vol = Volume(center=reg.center,size=reg.size)
            ax = plot_volume(sim,ax,vol,output_plane,plotting_parameters=monitor_parameters,label=label)
    return ax

def plot_fields(sim,ax=None,fields=None,output_plane=None,field_parameters=None):
    if not sim._is_initialized:
        sim.init_sim()

    if fields is None:
        return ax

    field_parameters = default_field_parameters if field_parameters is None else dict(default_field_parameters, **field_parameters)

    # user specifies a field component
    if fields in [mp.Ex, mp.Ey, mp.Ez, mp.Hx, mp.Hy, mp.Hz]:
        # Get domain measurements
        sim_center, sim_size = get_2D_dimensions(sim,output_plane)

        xmin = sim_center.x - sim_size.x/2
        xmax = sim_center.x + sim_size.x/2
        ymin = sim_center.y - sim_size.y/2
        ymax = sim_center.y + sim_size.y/2
        zmin = sim_center.z - sim_size.z/2
        zmax = sim_center.z + sim_size.z/2

        center = Vector3(sim_center.x,sim_center.y,sim_center.z)
        cell_size = Vector3(sim_size.x,sim_size.y,sim_size.z)

        if sim_size.x == 0:
            # Plot y on x axis, z on y axis (YZ plane)
            extent = [ymin,ymax,zmin,zmax]
            xlabel = 'Y'
            ylabel = 'Z'
        elif sim_size.y == 0:
            # Plot x on x axis, z on y axis (XZ plane)
            extent = [xmin,xmax,zmin,zmax]
            xlabel = 'X'
            ylabel = 'Z'
        elif sim_size.z == 0:
            # Plot x on x axis, y on y axis (XY plane)
            extent = [xmin,xmax,ymin,ymax]
            xlabel = 'X'
            ylabel = 'Y'
        fields = sim.get_array(center=center, size=cell_size, component=fields)
    else:
        raise ValueError('Please specify a valid field component (mp.Ex, mp.Ey, ...')


    fields = field_parameters['post_process'](fields)

    # Either plot the field, or return the array
    if ax:
        if mp.am_master():
            ax.imshow(np.rot90(fields), extent=extent, **filter_dict(field_parameters,ax.imshow))
        return ax
    else:
        return np.rot90(fields)
    return ax

def plot2D(sim,ax=None, output_plane=None, fields=None, labels=False,
           eps_parameters=None,boundary_parameters=None,
           source_parameters=None,monitor_parameters=None,
           field_parameters=None, frequency=None,
           plot_eps_flag=True, plot_sources_flag=True,
           plot_monitors_flag=True, plot_boundaries_flag=True):

    # Initialize the simulation
    if sim.structure is None:
        sim.init_sim()
    # Ensure a figure axis exists
    if ax is None and mp.am_master():
        from matplotlib import pyplot as plt
        ax = plt.gca()
    # Determine a frequency to plot all epsilon
    if frequency is None:
        try:
            # Continuous sources
            frequency = sim.sources[0].frequency
        except:
            try:
                # Gaussian sources
                frequency = sim.sources[0].src.frequency
            except:
                try:
                    # Custom sources
                    frequency = sim.sources[0].src.center_frequency
                except:
                    # No sources
                    frequency = 0

    # validate the output plane to ensure proper 2D coordinates
    from meep.simulation import Volume
    sim_center, sim_size = get_2D_dimensions(sim,output_plane)
    output_plane = Volume(center=sim_center,size=sim_size)

    # Plot geometry
    if plot_eps_flag:
        ax = plot_eps(sim,ax,output_plane=output_plane,eps_parameters=eps_parameters,frequency=frequency)

    # Plot boundaries
    if plot_boundaries_flag:
        ax = plot_boundaries(sim,ax,output_plane=output_plane,boundary_parameters=boundary_parameters)

    # Plot sources
    if plot_sources_flag:
        ax = plot_sources(sim,ax,output_plane=output_plane,labels=labels,source_parameters=source_parameters)

    # Plot monitors
    if plot_monitors_flag:
        ax = plot_monitors(sim,ax,output_plane=output_plane,labels=labels,monitor_parameters=monitor_parameters)

    # Plot fields
    ax = plot_fields(sim,ax,fields,output_plane=output_plane,field_parameters=field_parameters)

    return ax

def plot3D(sim):
    from mayavi import mlab

    if not sim._is_initialized:
        sim.init_sim()

    if sim.dimensions < 3:
        raise ValueError("Simulation must have 3 dimensions to visualize 3D")

    eps_data = sim.get_epsilon()
    s = mlab.contour3d(eps_data, colormap="YlGnBu")
    return s

def visualize_chunks(sim):
    if sim.structure is None:
        sim.init_sim()

    import matplotlib.pyplot as plt
    import matplotlib.cm
    import matplotlib.colors

    if sim.structure.gv.dim == 2:
        from mpl_toolkits.mplot3d import Axes3D
        from mpl_toolkits.mplot3d.art3d import Poly3DCollection
    else:
        from matplotlib.collections import PolyCollection

    vols = sim.structure.get_chunk_volumes()
    owners = sim.structure.get_chunk_owners()

    def plot_box(box, proc, fig, ax):
        if sim.structure.gv.dim == 2:
            low = mp.Vector3(box.low.x, box.low.y, box.low.z)
            high = mp.Vector3(box.high.x, box.high.y, box.high.z)
            points = [low, high]

            x_len = mp.Vector3(high.x) - mp.Vector3(low.x)
            y_len = mp.Vector3(y=high.y) - mp.Vector3(y=low.y)
            xy_len = mp.Vector3(high.x, high.y) - mp.Vector3(low.x, low.y)

            points += [low + x_len]
            points += [low + y_len]
            points += [low + xy_len]
            points += [high - x_len]
            points += [high - y_len]
            points += [high - xy_len]
            points = np.array([np.array(v) for v in points])

            edges = [
                [points[0], points[2], points[4], points[3]],
                [points[1], points[5], points[7], points[6]],
                [points[0], points[3], points[5], points[7]],
                [points[1], points[4], points[2], points[6]],
                [points[3], points[4], points[1], points[5]],
                [points[0], points[7], points[6], points[2]]
            ]

            faces = Poly3DCollection(edges, linewidths=1, edgecolors='k')
            color_with_alpha = matplotlib.colors.to_rgba(chunk_colors[proc], alpha=0.2)
            faces.set_facecolor(color_with_alpha)
            ax.add_collection3d(faces)

            # Plot the points themselves to force the scaling of the axes
            ax.scatter(points[:, 0], points[:, 1], points[:, 2], s=0)
        else:
            low = mp.Vector3(box.low.x, box.low.y)
            high = mp.Vector3(box.high.x, box.high.y)
            points = [low, high]

            x_len = mp.Vector3(high.x) - mp.Vector3(low.x)
            y_len = mp.Vector3(y=high.y) - mp.Vector3(y=low.y)

            points += [low + x_len]
            points += [low + y_len]
            points = np.array([np.array(v)[:-1] for v in points])

            edges = [
                [points[0], points[2], points[1], points[3]]
            ]

            faces = PolyCollection(edges, linewidths=1, edgecolors='k')
            color_with_alpha = matplotlib.colors.to_rgba(chunk_colors[proc])
            faces.set_facecolor(color_with_alpha)
            ax.add_collection(faces)

            # Plot the points themselves to force the scaling of the axes
            ax.scatter(points[:, 0], points[:, 1], s=0)

    if mp.am_master():
        fig = plt.figure()
        ax = fig.add_subplot(111, projection='3d' if sim.structure.gv.dim == 2 else None)
        chunk_colors = matplotlib.cm.rainbow(np.linspace(0, 1, mp.count_processors()))

        for i, v in enumerate(vols):
            plot_box(mp.gv2box(v.surroundings()), owners[i], fig, ax)

        ax.set_xlabel('x')
        ax.set_ylabel('y')
        ax.set_aspect('equal')

        cell_box = mp.gv2box(sim.structure.gv.surroundings())
        if sim.structure.gv.dim == 2:
            ax.set_xlim3d(left=cell_box.low.x,right=cell_box.high.x)
            ax.set_ylim3d(bottom=cell_box.low.y,top=cell_box.high.y)
            ax.set_zlim3d(bottom=cell_box.low.z,top=cell_box.high.z)
            ax.set_zlabel('z')
        else:
            ax.set_xlim(left=cell_box.low.x,right=cell_box.high.x)
            ax.set_ylim(bottom=cell_box.low.y,top=cell_box.high.y)

        plt.tight_layout()
        plt.show()

# ------------------------------------------------------- #
# Animate2D
# ------------------------------------------------------- #
# An extensive run function used to visualize the fields
# of a 2D simulation after every specified time step.
# ------------------------------------------------------- #
# Required arguments
# sim ................. [Simulation object]
# fields .............. [mp.Ex, mp.Ey, ..., mp. Hz]
# ------------------------------------------------------- #
# Optional arguments
# f ................... [matplotlib figure object]
# realtime ............ [bool] Update plot in each step
# normalize ........... [bool] saves fields to normalize
#                       after simulation ends.
# plot_modifiers ...... [list] additional functions to
#                       modify plot
# customization_args .. [dict] other customization args
#                       to pass to plot2D()

class Animate2D(object):
    """
    A class used to record the fields during timestepping (i.e., a [`run`](#run-functions)
    function). The object is initialized prior to timestepping by specifying the
    simulation object and the field component. The object can then be passed to any
    [step-function modifier](#step-function-modifiers). For example, one can record the
    E<sub>z</sub> fields at every one time unit using:

    ```py
    animate = mp.Animate2D(sim,
                           fields=mp.Ez,
                           realtime=True,
                           field_parameters={'alpha':0.8, 'cmap':'RdBu', 'interpolation':'none'},
                           boundary_parameters={'hatch':'o', 'linewidth':1.5, 'facecolor':'y', 'edgecolor':'b', 'alpha':0.3})

    sim.run(mp.at_every(1,animate),until=25)
    ```

    By default, the object saves each frame as a PNG image into memory (not disk). This is
    typically more memory efficient than storing the actual fields. If the user sets the
    `normalize` argument, then the object will save the actual field information as a
    NumPy array to be normalized for post processing. The fields of a figure can also be
    updated in realtime by setting the `realtime` flag. This does not work for
    IPython/Jupyter notebooks, however.

    Once the simulation is run, the animation can be output as an interactive JSHTML
    object, an mp4, or a GIF.

    Multiple `Animate2D` objects can be initialized and passed to the run function to
    track different volume locations (using `mp.in_volume`) or field components.
    """
    def __init__(self, sim, fields, f=None, realtime=False, normalize=False,
                 plot_modifiers=None, **customization_args):
        """
        Construct an `Animate2D` object.

        + **`sim`** — Simulation object.

        + **`fields`** — Field component to record at each time instant.

        + **`f=None`** — Optional `matplotlib` figure object that the routine will update
          on each call. If not supplied, then a new one will be created upon
          initialization.

        + **`realtime=False`** — Whether or not to update a figure window in realtime as
          the simulation progresses. Disabled by default. Not compatible with
          IPython/Jupyter notebooks.

        + **`normalize=False`** — Records fields at each time step in memory in a NumPy
          array and then normalizes the result by dividing by the maximum field value at a
          single point in the cell over all the time snapshots.

        + **`plot_modifiers=None`** — A list of functions that can modify the figure's
          `axis` object. Each function modifier accepts a single argument, an `axis`
          object, and must return that same axis object. The following modifier changes
          the `xlabel`:

        ```py
          def mod1(ax):
              ax.set_xlabel('Testing')
              return ax

          plot_modifiers = [mod1]
        ```

        + **`**customization_args`** — Customization keyword arguments passed to
          `plot2D()` (i.e. `labels`, `eps_parameters`, `boundary_parameters`, etc.)
        """

        self.fields = fields

        if f:
            self.f = f
            self.ax = self.f.gca()
        elif mp.am_master():
            from matplotlib import pyplot as plt
            self.f = plt.figure()
            self.ax = self.f.gca()
        else:
            self.f = None
            self.ax = None

        self.realtime = realtime
        self.normalize = normalize
        self.plot_modifiers = plot_modifiers
        self.customization_args = customization_args


        self.cumulative_fields = []
        self._saved_frames = []

        self.frame_format = 'png' # format in which each frame is saved in memory
        self.codec = 'h264' # encoding of mp4 video
        self.default_mode = 'loop' # html5 video control mode

        self.init = False

        # Needed for step functions
        self.__code__ = namedtuple('gna_hack',['co_argcount'])
        self.__code__.co_argcount=2

    def __call__(self,sim,todo):
        from matplotlib import pyplot as plt

        if todo == 'step':
            # Initialize the plot
            if not self.init:
                filtered_plot2D = filter_dict(self.customization_args, plot2D)
                ax = sim.plot2D(ax=self.ax,fields=self.fields,**filtered_plot2D)
                # Run the plot modifier functions
                if self.plot_modifiers:
                    for k in range(len(self.plot_modifiers)):
                        ax = self.plot_modifiers[k](self.ax)
                # Store the fields
                if mp.am_master():
                    fields = ax.images[-1].get_array()
                    self.ax = ax
                    self.w, self.h = self.f.get_size_inches()
                self.init = True
            else:
                # Update the plot
                filtered_plot_fields= filter_dict(self.customization_args, plot_fields)
                fields = sim.plot_fields(fields=self.fields,**filtered_plot_fields)
                if mp.am_master():
                    self.ax.images[-1].set_data(fields)
                    self.ax.images[-1].set_clim(vmin=0.8*np.min(fields), vmax=0.8*np.max(fields))

            if self.realtime and mp.am_master():
                # Redraw the current figure if requested
                plt.pause(0.05)

            if self.normalize and mp.am_master():
                # Save fields as a numpy array to be normalized
                # and saved later.
                self.cumulative_fields.append(fields)
            elif mp.am_master():
                # Capture figure as a png, but store the png in memory
                # to avoid writing to disk.
                self.grab_frame()
            return
        elif todo == 'finish':

            # Normalize the frames, if requested, and export
            if self.normalize and mp.am_master():
                if mp.verbosity.meep > 0:
                    print("Normalizing field data...")
                fields = np.array(self.cumulative_fields) / np.max(np.abs(self.cumulative_fields),axis=(0,1,2))
                for k in range(len(self.cumulative_fields)):
                    self.ax.images[-1].set_data(fields[k,:,:])
                    self.ax.images[-1].set_clim(vmin=-0.8, vmax=0.8)
                    self.grab_frame()

            return

    @property
    def frame_size(self):
        # A tuple ``(width, height)`` in pixels of a movie frame.
        # modified from matplotlib library
        w, h = self.f.get_size_inches()
        return int(w * self.f.dpi), int(h * self.f.dpi)

    def grab_frame(self):
        # Saves the figures frame to memory.
        # modified from matplotlib library
        from io import BytesIO
        bin_data = BytesIO()
        self.f.savefig(bin_data, format=self.frame_format)
        #imgdata64 = base64.encodebytes(bin_data.getvalue()).decode('ascii')
        self._saved_frames.append(bin_data.getvalue())

    def _embedded_frames(self, frame_list, frame_format):
        # converts frame data stored in memory to html5 friendly format
        # frame_list should be a list of base64-encoded png files
        # modified from matplotlib
        import base64
        template = '  frames[{0}] = "data:image/{1};base64,{2}"\n'
        return "\n" + "".join(
            template.format(i, frame_format, base64.encodebytes(frame_data).decode('ascii').replace('\n', '\\\n'))
            for i, frame_data in enumerate(frame_list))

    def to_jshtml(self,fps):
        """
        Outputs an interactable JSHTML animation object that is embeddable in Jupyter
        notebooks. The object is packaged with controls to manipulate the video's
        playback. User must specify a frame rate `fps` in frames per second.
        """
        # Exports a javascript enabled html object that is
        # ready for jupyter notebook embedding.
        # modified from matplotlib/animation.py code.

        # Only works with Python3 and matplotlib > 3.1.0
        from distutils.version import LooseVersion
        import matplotlib
        if LooseVersion(matplotlib.__version__) < LooseVersion("3.1.0"):
            print('-------------------------------')
            print('Warning: JSHTML output is not supported with your current matplotlib build. Consider upgrading to 3.1.0+')
            print('-------------------------------')
            return
        if mp.am_master():
            from uuid import uuid4
            from matplotlib._animation_data import (DISPLAY_TEMPLATE, INCLUDED_FRAMES, JS_INCLUDE, STYLE_INCLUDE)

            # save the frames to an html file
            fill_frames = self._embedded_frames(self._saved_frames, self.frame_format)
            Nframes = len(self._saved_frames)
            mode_dict = dict(once_checked='',
                            loop_checked='',
                            reflect_checked='')
            mode_dict[self.default_mode + '_checked'] = 'checked'

            interval = 1000 // fps

            html_string = ""
            html_string += JS_INCLUDE
            html_string += STYLE_INCLUDE
            html_string += DISPLAY_TEMPLATE.format(id=uuid4().hex,
                                                Nframes=Nframes,
                                                fill_frames=fill_frames,
                                                interval=interval,
                                                **mode_dict)
            return JS_Animation(html_string)

    def to_gif(self,fps,filename):
        """
        Generates and outputs a GIF file of the animation with the filename, `filename`,
        and the frame rate, `fps`. Note that GIFs are significantly larger than mp4 videos
        since they don't use any compression. Artifacts are also common because the GIF
        format only supports 256 colors from a _predefined_ color palette. Requires
        `ffmpeg`.
        """
        # Exports a gif of the recorded animation
        # requires ffmpeg to be installed
        # modified from the matplotlib library
        if mp.am_master():
            from subprocess import Popen, PIPE
            from io import TextIOWrapper, BytesIO
            FFMPEG_BIN = 'ffmpeg'
            command = [FFMPEG_BIN,
                        '-f', 'image2pipe', # force piping of rawvideo
                        '-vcodec', self.frame_format, # raw input codec
                        '-s', '%dx%d' % (self.frame_size),
                        '-r', str(fps), # frame rate in frames per second
                        '-i', 'pipe:', # The input comes from a pipe
                        '-vcodec', 'gif', # output gif format
                        '-r', str(fps), # frame rate in frames per second
                        '-y',
                        '-vf', 'pad=width=ceil(iw/2)*2:height=ceil(ih/2)*2',
                        '-an', filename  # output filename
            ]
            if mp.verbosity.meep > 0:
                print("Generating GIF...")
            proc = Popen(command, stdin=PIPE, stdout=PIPE, stderr=PIPE)
            for i in range(len(self._saved_frames)):
                proc.stdin.write(self._saved_frames[i])
            out, err = proc.communicate() # pipe in images
            proc.stdin.close()
            proc.wait()
        return

    def to_mp4(self,fps,filename):
        """
        Generates and outputs an mp4 video file of the animation with the filename,
        `filename`, and the frame rate, `fps`. Default encoding is h264 with yuv420p
        format. Requires `ffmpeg`.
        """
        # Exports an mp4 of the recorded animation
        # requires ffmpeg to be installed
        # modified from the matplotlib library
        if mp.am_master():
            from subprocess import Popen, PIPE
            from io import TextIOWrapper, BytesIO
            FFMPEG_BIN = 'ffmpeg'
            command = [FFMPEG_BIN,
                        '-f', 'image2pipe', # force piping of rawvideo
                        '-vcodec', self.frame_format, # raw input codec
                        '-s', '%dx%d' % (self.frame_size),
                        #'-pix_fmt', self.frame_format,
                        '-r', str(fps), # frame rate in frames per second
                        '-i', 'pipe:', # The input comes from a pipe
                        '-vcodec', self.codec, # output mp4 format
                        '-pix_fmt','yuv420p',
                        '-r', str(fps), # frame rate in frames per second
                        '-y',
                        '-vf', 'pad=width=ceil(iw/2)*2:height=ceil(ih/2)*2',
                        '-an', filename  # output filename
            ]
            if mp.verbosity.meep > 0:
                print("Generating MP4...")
            proc = Popen(command, stdin=PIPE, stdout=PIPE, stderr=PIPE)
            for i in range(len(self._saved_frames)):
                proc.stdin.write(self._saved_frames[i])
            out, err = proc.communicate() # pipe in images
            proc.stdin.close()
            proc.wait()
        return

    def reset(self):
        self.cumulative_fields = []
        self.ax = None
        self.f = None

    def set_figure(self,f):
        self.f = f

# ------------------------------------------------------- #
# JS_Animation
# ------------------------------------------------------- #
# A helper class used to make jshtml animations embed
# seamlessly within Jupyter notebooks.

class JS_Animation():
    def __init__(self,jshtml):
        self.jshtml = jshtml
    def _repr_html_(self):
        return self.jshtml
    def get_jshtml(self):
        return self.jshtml