maps.py

# -*- coding: utf-8 -*-
"""
Module for basemap related plotting in ObsPy.

:copyright:
    The ObsPy Development Team (devs@obspy.org)
:license:
    GNU Lesser General Public License, Version 3
    (https://www.gnu.org/copyleft/lesser.html)
"""
from __future__ import (absolute_import, division, print_function,
                        unicode_literals)
from future.builtins import *  # NOQA @UnusedWildImport
from future.utils import native_str

import datetime
import warnings

import numpy as np
from matplotlib.colorbar import Colorbar
from matplotlib.dates import AutoDateFormatter, AutoDateLocator, date2num
from matplotlib import patheffects
from matplotlib.ticker import (FormatStrFormatter, Formatter, FuncFormatter,
                               MaxNLocator)

from obspy import UTCDateTime
from obspy.core.util import (BASEMAP_VERSION, CARTOPY_VERSION,
                             MATPLOTLIB_VERSION, PROJ4_VERSION)
from obspy.geodetics.base import mean_longitude


if BASEMAP_VERSION:
    from mpl_toolkits.basemap import Basemap
    HAS_BASEMAP = True
    if BASEMAP_VERSION < [1, 0, 4]:
        warnings.warn("All basemap version < 1.0.4 contain a serious bug "
                      "when rendering countries and continents. ObsPy will "
                      "still work but the maps might be wrong. Please update "
                      "your basemap installation.")
    if PROJ4_VERSION and PROJ4_VERSION[0] == 5:
        msg = (
            "basemap/pyproj with proj4 version >= 5 has a bug that results in "
            "inverted map axes. Your maps may be wrong. Please use another "
            "version of proj4, or use cartopy. "
            "See https://github.com/matplotlib/basemap/issues/443")
        warnings.warn(msg)
    if MATPLOTLIB_VERSION == [3, 0, 1] and BASEMAP_VERSION >= [1, 1, 0]:
        msg = (
            "basemap and matplotlib version 3.0.1 have compataibilty issues, "
            "please change your matplotlib version. "
            "See https://github.com/matplotlib/basemap/issues/435")
        warnings.warn(msg)
else:
    HAS_BASEMAP = False

if CARTOPY_VERSION and CARTOPY_VERSION >= [0, 12, 0]:
    import cartopy.crs as ccrs
    import cartopy.feature as cfeature
    HAS_CARTOPY = True
else:
    HAS_CARTOPY = False


if not HAS_BASEMAP and not HAS_CARTOPY:
    msg = ("Neither basemap nor cartopy installed, map plots will not work.")
    warnings.warn(msg)


_BASEMAP_RESOLUTIONS = {
    '110m': 'l',
    '50m': 'i',
    '10m': 'f',
    'c': 'c',
    'l': 'l',
    'i': 'i',
    'h': 'h',
    'f': 'f',
}

_CARTOPY_RESOLUTIONS = {
    'c': '110m',
    'l': '110m',
    'i': '50m',
    'h': '50m',
    'f': '10m',
    '110m': '110m',
    '50m': '50m',
    '10m': '10m',
}

if HAS_CARTOPY:
    _CARTOPY_FEATURES = {
        '110m': (cfeature.BORDERS, cfeature.LAND, cfeature.OCEAN),
    }


def plot_basemap(lons, lats, size, color, labels=None, projection='global',
                 resolution='l', continent_fill_color='0.8',
                 water_fill_color='1.0', colormap=None, colorbar=None,
                 marker="o", title=None, colorbar_ticklabel_format=None,
                 show=True, fig=None, **kwargs):  # @UnusedVariable
    """
    Creates a basemap plot with a data point scatter plot.

    :type lons: list/tuple of floats
    :param lons: Longitudes of the data points.
    :type lats: list/tuple of floats
    :param lats: Latitudes of the data points.
    :type size: float or list/tuple of floats
    :param size: Size of the individual points in the scatter plot.
    :type color: list/tuple of floats (or objects that can be
        converted to floats, like e.g.
        :class:`~obspy.core.utcdatetime.UTCDateTime`)
    :param color: Color information of the individual data points to be
        used in the specified color map (e.g. origin depths,
        origin times).
    :type labels: list/tuple of str
    :param labels: Annotations for the individual data points.
    :type projection: str, optional
    :param projection: The map projection.
        Currently supported are:

            * ``"global"`` (Will plot the whole world.)
            * ``"ortho"`` (Will center around the mean lat/long.)
            * ``"local"`` (Will plot around local events)

        Defaults to "global".
    :type resolution: str, optional
    :param resolution: Resolution of the boundary database to use. Will be
        based directly to the basemap module. Possible values are:

            * ``"c"`` (crude)
            * ``"l"`` (low)
            * ``"i"`` (intermediate)
            * ``"h"`` (high)
            * ``"f"`` (full)

        Defaults to ``"l"``. For compatibility, you may also specify any of the
        Cartopy resolutions defined in :func:`plot_cartopy`.
    :type continent_fill_color: Valid matplotlib color, optional
    :param continent_fill_color:  Color of the continents. Defaults to
        ``"0.9"`` which is a light gray.
    :type water_fill_color: Valid matplotlib color, optional
    :param water_fill_color: Color of all water bodies.
        Defaults to ``"white"``.
    :type colormap: str, any matplotlib colormap, optional
    :param colormap: The colormap for color-coding the events as provided
        in `color` kwarg.
        The event with the smallest `color` property will have the
        color of one end of the colormap and the event with the highest
        `color` property the color of the other end with all other events
        in between.
        Defaults to None which will use the default matplotlib colormap.
    :type colorbar: bool, optional
    :param colorbar: When left `None`, a colorbar is plotted if more than one
        object is plotted. Using `True`/`False` the colorbar can be forced
        on/off.
    :type title: str
    :param title: Title above plot.
    :type colorbar_ticklabel_format: str or function or
        subclass of :class:`matplotlib.ticker.Formatter`
    :param colorbar_ticklabel_format: Format string or Formatter used to format
        colorbar tick labels.
    :type show: bool
    :param show: Whether to show the figure after plotting or not. Can be used
        to do further customization of the plot before showing it.
    :type fig: :class:`matplotlib.figure.Figure`
    :param fig: Figure instance to reuse, returned from a previous
        :func:`plot_basemap` call. If a previous basemap plot is reused, any
        kwargs regarding the basemap plot setup will be ignored (i.e.
        `projection`, `resolution`, `continent_fill_color`,
        `water_fill_color`). Note that multiple plots using colorbars likely
        are problematic, but e.g. one station plot (without colorbar) and one
        event plot (with colorbar) together should work well.
    """
    import matplotlib.pyplot as plt

    if any([isinstance(c, (datetime.datetime, UTCDateTime)) for c in color]):
        datetimeplot = True
        color = [
            (np.isfinite(float(t)) and
             date2num(getattr(t, 'datetime', t)) or
             np.nan)
            for t in color]
    else:
        datetimeplot = False

    # The colorbar should only be plotted if more then one event is
    # present.
    if colorbar is None:
        if len(lons) > 1 and hasattr(color, "__len__") and \
                not isinstance(color, (str, native_str)):
            colorbar = True
        else:
            colorbar = False

    if fig is None:
        fig = plt.figure()

        if projection == "local":
            ax_x0, ax_width = 0.10, 0.80
        elif projection == "global":
            ax_x0, ax_width = 0.01, 0.98
        else:
            ax_x0, ax_width = 0.05, 0.90

        if colorbar:
            map_ax = fig.add_axes([ax_x0, 0.13, ax_width, 0.77])
            cm_ax = fig.add_axes([ax_x0, 0.05, ax_width, 0.05])
        else:
            ax_y0, ax_height = 0.05, 0.85
            if projection == "local":
                ax_y0 += 0.05
                ax_height -= 0.05
            map_ax = fig.add_axes([ax_x0, ax_y0, ax_width, ax_height])
        bmap = None

    else:
        error_message_suffix = (
            ". Please provide a figure object from a previous call to the "
            ".plot() method of e.g. an Inventory or Catalog object.")
        try:
            map_ax = fig.axes[0]
        except IndexError as e:
            e.args = tuple([e.args[0] + error_message_suffix] +
                           list(e.args[1:]))
            raise
        try:
            bmap = fig.bmap
        except AttributeError as e:
            e.args = tuple([e.args[0] + error_message_suffix] +
                           list(e.args[1:]))
            raise

    # basemap plots will break with basemap 1.1.0 together with matplotlib
    # >=2.3 (see matplotlib/basemap#382) so only thing we can do is show a
    # nicer message.
    # XXX can be removed maybe a year or so after basemap
    # 1.1.1 or 1.2.0 is released
    try:
        from matplotlib.cbook import MatplotlibDeprecationWarning
    except ImportError:
        # matplotlib 1.2.0 does not have that warning class yet
        # XXX can be removed when minimum matplotlib version gets bumped to
        # XXX 1.3.0
        category = {}
    else:
        category = {'category': MatplotlibDeprecationWarning}
    try:
        with warnings.catch_warnings():
            warnings.filterwarnings(
                'ignore', message='The axesPatch function was deprecated '
                'in version 2.1. Use Axes.patch instead.',
                module='.*basemap.*', **category)
            scatter = _plot_basemap_into_axes(
                ax=map_ax, lons=lons, lats=lats, size=size, color=color,
                bmap=bmap, labels=labels, projection=projection,
                resolution=resolution,
                continent_fill_color=continent_fill_color,
                water_fill_color=water_fill_color, colormap=colormap,
                marker=marker, title="", adjust_aspect_to_colorbar=colorbar,
                **kwargs)
    except AttributeError as e:
        if 'axesPatch' not in str(e):
            raise
        msg = ('Encountered a problem doing the basemap plot due to a known '
               'issue of matplotlib >=2.3 together with basemap <=1.1.0 (see '
               'https://github.com/matplotlib/basemap/issues/382). Please '
               'update basemap to a version >1.1.0 if available or downgrade '
               'matplotlib to a version <2.3.')
        raise Exception(msg)

    if title:
        plt.suptitle(title)

    if colorbar:
        if colorbar_ticklabel_format is not None:
            if isinstance(colorbar_ticklabel_format, (str, native_str)):
                formatter = FormatStrFormatter(colorbar_ticklabel_format)
            elif hasattr(colorbar_ticklabel_format, '__call__'):
                formatter = FuncFormatter(colorbar_ticklabel_format)
            elif isinstance(colorbar_ticklabel_format, Formatter):
                formatter = colorbar_ticklabel_format
            locator = MaxNLocator(5)
        else:
            if datetimeplot:
                locator = AutoDateLocator()
                formatter = AutoDateFormatter(locator)
                # Compat with old matplotlib versions.
                if hasattr(formatter, "scaled"):
                    formatter.scaled[1 / (24. * 60.)] = '%H:%M:%S'
            else:
                locator = None
                formatter = None

        # normal case: axes for colorbar was set up in this routine
        if "cm_ax" in locals():
            cb_kwargs = {"cax": cm_ax}
        # unusual case: reusing a plot that has no colorbar set up previously
        else:
            cb_kwargs = {"ax": map_ax}

        cb = fig.colorbar(
            mappable=scatter, cmap=colormap, orientation='horizontal',
            ticks=locator, format=formatter, **cb_kwargs)
        # Compat with old matplotlib versions.
        if hasattr(cb, "update_ticks"):
            cb.update_ticks()

    if show:
        plt.show()

    return fig


def _plot_basemap_into_axes(
        ax, lons, lats, size, color, bmap=None, labels=None,
        projection='global', resolution='l', continent_fill_color='0.8',
        water_fill_color='1.0', colormap=None, marker="o", title=None,
        adjust_aspect_to_colorbar=False, **kwargs):  # @UnusedVariable
    """
    Creates a (or adds to existing) basemap plot with a data point scatter
    plot in given axes.

    See :func:`plot_basemap` for details on most args/kwargs.

    :type ax: :class:`matplotlib.axes.Axes`
    :param ax: Existing matplotlib axes instance, optionally with previous
        basemap plot (see `bmap` kwarg).
    :type bmap: :class:`mpl_toolkits.basemap.Basemap`
    :param bmap: Basemap instance in provided matplotlib Axes `ax` to reuse. If
        specified, any kwargs regarding the basemap plot setup will be ignored
        (i.e.  `projection`, `resolution`, `continent_fill_color`,
        `water_fill_color`).
    :rtype: :class:`matplotlib.collections.PathCollection`
    :returns: Matplotlib path collection (e.g. to reuse for colorbars).
    """
    fig = ax.figure
    if bmap is None:

        if projection == 'global':
            bmap = Basemap(projection='moll', lon_0=round(np.mean(lons), 4),
                           resolution=_BASEMAP_RESOLUTIONS[resolution],
                           ax=ax)
        elif projection == 'ortho':
            bmap = Basemap(projection='ortho',
                           resolution=_BASEMAP_RESOLUTIONS[resolution],
                           lat_0=round(np.mean(lats), 4),
                           lon_0=round(mean_longitude(lons), 4), ax=ax)
        elif projection == 'local':
            if min(lons) < -150 and max(lons) > 150:
                max_lons = max(np.array(lons) % 360)
                min_lons = min(np.array(lons) % 360)
            else:
                max_lons = max(lons)
                min_lons = min(lons)
            lat_0 = max(lats) / 2. + min(lats) / 2.
            lon_0 = max_lons / 2. + min_lons / 2.
            if lon_0 > 180:
                lon_0 -= 360
            deg2m_lat = 2 * np.pi * 6371 * 1000 / 360
            deg2m_lon = deg2m_lat * np.cos(lat_0 / 180 * np.pi)
            if len(lats) > 1:
                height = (max(lats) - min(lats)) * deg2m_lat
                width = (max_lons - min_lons) * deg2m_lon
                margin = 0.2 * (width + height)
                height += margin
                width += margin
            else:
                height = 2.0 * deg2m_lat
                width = 5.0 * deg2m_lon
            # do intelligent aspect calculation for local projection
            # adjust to figure dimensions
            w, h = fig.get_size_inches()
            ax_bbox = ax.get_position()
            aspect = (w * ax_bbox.width) / (h * ax_bbox.height)
            if adjust_aspect_to_colorbar:
                aspect *= 1.2
            if width / height < aspect:
                width = height * aspect
            else:
                height = width / aspect

            bmap = Basemap(projection='aea',
                           resolution=_BASEMAP_RESOLUTIONS[resolution],
                           lat_0=round(lat_0, 4),
                           lon_0=round(lon_0, 4),
                           width=width, height=height, ax=ax)
            # not most elegant way to calculate some round lats/lons

            def linspace2(val1, val2, n):
                """
                returns around n 'nice' values between val1 and val2
                """
                dval = val2 - val1
                round_pos = int(round(-np.log10(1. * dval / n)))
                # Fake negative rounding as not supported by future as of now.
                if round_pos < 0:
                    factor = 10 ** (abs(round_pos))
                    delta = round(2. * dval / n / factor) * factor / 2
                else:
                    delta = round(2. * dval / n, round_pos) / 2
                new_val1 = np.ceil(val1 / delta) * delta
                new_val2 = np.floor(val2 / delta) * delta
                n = int((new_val2 - new_val1) / delta + 1)
                return np.linspace(new_val1, new_val2, n)

            n_1 = int(np.ceil(height / max(width, height) * 8))
            n_2 = int(np.ceil(width / max(width, height) * 8))
            parallels = linspace2(lat_0 - height / 2 / deg2m_lat,
                                  lat_0 + height / 2 / deg2m_lat, n_1)

            # Old basemap versions have problems with non-integer parallels.
            try:
                bmap.drawparallels(parallels, labels=[0, 1, 1, 0])
            except KeyError:
                parallels = sorted(list(set(map(int, parallels))))
                bmap.drawparallels(parallels, labels=[0, 1, 1, 0])

            if min(lons) < -150 and max(lons) > 150:
                lon_0 %= 360
            meridians = linspace2(lon_0 - width / 2 / deg2m_lon,
                                  lon_0 + width / 2 / deg2m_lon, n_2)
            meridians[meridians > 180] -= 360
            bmap.drawmeridians(meridians, labels=[1, 0, 0, 1])
        else:
            msg = "Projection '%s' not supported." % projection
            raise ValueError(msg)

        # draw coast lines, country boundaries, fill continents.
        if MATPLOTLIB_VERSION >= [2, 0, 0]:
            ax.set_facecolor(water_fill_color)
        else:
            ax.set_axis_bgcolor(water_fill_color)
        bmap.drawcoastlines(color="0.4")
        bmap.drawcountries(color="0.75")
        bmap.fillcontinents(color=continent_fill_color,
                            lake_color=water_fill_color)
        # draw the edge of the bmap projection region (the projection limb)
        bmap.drawmapboundary(fill_color=water_fill_color)
        # draw lat/lon grid lines every 30 degrees.
        bmap.drawmeridians(np.arange(-180, 180, 30))
        bmap.drawparallels(np.arange(-90, 90, 30))
        fig.bmap = bmap

    # compute the native bmap projection coordinates for events.
    x, y = bmap(lons, lats)
    # plot labels
    if labels:
        if 100 > len(lons) > 1:
            for name, xpt, ypt, _colorpt in zip(labels, x, y, color):
                # Check if the point can actually be seen with the current bmap
                # projection. The bmap object will set the coordinates to very
                # large values if it cannot project a point.
                if xpt > 1e25:
                    continue
                ax.text(xpt, ypt, name, weight="heavy",
                        color="k", zorder=100,
                        path_effects=[
                            patheffects.withStroke(linewidth=3,
                                                   foreground="white")])
        elif len(lons) == 1:
            ax.text(x[0], y[0], labels[0], weight="heavy", color="k",
                    path_effects=[
                        patheffects.withStroke(linewidth=3,
                                               foreground="white")])

    # scatter plot is removing valid x/y points with invalid color value,
    # so we plot those points separately.
    try:
        nan_points = np.isnan(np.array(color, dtype=np.float))
    except ValueError:
        # `color' was not a list of values, but a list of colors.
        pass
    else:
        if nan_points.any():
            x_ = np.array(x)[nan_points]
            y_ = np.array(y)[nan_points]
            size_ = np.array(size)[nan_points]
            bmap.scatter(x_, y_, marker=marker, s=size_, c="0.3",
                         zorder=10, cmap=None)
    scatter = bmap.scatter(x, y, marker=marker, s=size, c=color, zorder=10,
                           cmap=colormap)

    if title:
        ax.set_title(title)

    return scatter


def plot_cartopy(lons, lats, size, color, labels=None, projection='global',
                 resolution='110m', continent_fill_color='0.8',
                 water_fill_color='1.0', colormap=None, colorbar=None,
                 marker="o", title=None, colorbar_ticklabel_format=None,
                 show=True, proj_kwargs=None, **kwargs):  # @UnusedVariable
    """
    Creates a Cartopy plot with a data point scatter plot.

    :type lons: list/tuple of floats
    :param lons: Longitudes of the data points.
    :type lats: list/tuple of floats
    :param lats: Latitudes of the data points.
    :type size: float or list/tuple of floats
    :param size: Size of the individual points in the scatter plot.
    :type color: list/tuple of floats (or objects that can be
        converted to floats, like e.g.
        :class:`~obspy.core.utcdatetime.UTCDateTime`)
    :param color: Color information of the individual data points to be
        used in the specified color map (e.g. origin depths,
        origin times).
    :type labels: list/tuple of str
    :param labels: Annotations for the individual data points.
    :type projection: str, optional
    :param projection: The map projection.
        Currently supported are:

            * ``"global"`` (Will plot the whole world using
              :class:`~cartopy.crs.Mollweide`.)
            * ``"ortho"`` (Will center around the mean lat/long using
              :class:`~cartopy.crs.Orthographic`.)
            * ``"local"`` (Will plot around local events using
              :class:`~cartopy.crs.AlbersEqualArea`.)
            * Any other Cartopy :class:`~cartopy.crs.Projection`. An instance
              of this class will be created using the supplied ``proj_kwargs``.

        Defaults to "global"
    :type resolution: str, optional
    :param resolution: Resolution of the boundary database to use. Will be
        passed directly to the Cartopy module. Possible values are:

            * ``"110m"``
            * ``"50m"``
            * ``"10m"``

        Defaults to ``"110m"``. For compatibility, you may also specify any of
        the Basemap resolutions defined in :func:`plot_basemap`.
    :type continent_fill_color: Valid matplotlib color, optional
    :param continent_fill_color:  Color of the continents. Defaults to
        ``"0.9"`` which is a light gray.
    :type water_fill_color: Valid matplotlib color, optional
    :param water_fill_color: Color of all water bodies.
        Defaults to ``"white"``.
    :type colormap: str, any matplotlib colormap, optional
    :param colormap: The colormap for color-coding the events as provided
        in `color` kwarg.
        The event with the smallest `color` property will have the
        color of one end of the colormap and the event with the highest
        `color` property the color of the other end with all other events
        in between.
        Defaults to None which will use the default matplotlib colormap.
    :type colorbar: bool, optional
    :param colorbar: When left `None`, a colorbar is plotted if more than one
        object is plotted. Using `True`/`False` the colorbar can be forced
        on/off.
    :type title: str
    :param title: Title above plot.
    :type colorbar_ticklabel_format: str or function or
        subclass of :class:`matplotlib.ticker.Formatter`
    :param colorbar_ticklabel_format: Format string or Formatter used to format
        colorbar tick labels.
    :type show: bool
    :param show: Whether to show the figure after plotting or not. Can be used
        to do further customization of the plot before showing it.
    :type proj_kwargs: dict
    :param proj_kwargs: Keyword arguments to pass to the Cartopy
        :class:`~cartopy.ccrs.Projection`. In this dictionary, you may specify
        ``central_longitude='auto'`` or ``central_latitude='auto'`` to have
        this function calculate the latitude or longitude as it would for other
        projections. Some arguments may be ignored if you choose one of the
        built-in ``projection`` choices.
    """
    import matplotlib.pyplot as plt

    if isinstance(color[0], (datetime.datetime, UTCDateTime)):
        datetimeplot = True
        color = [date2num(getattr(t, 'datetime', t)) for t in color]
    else:
        datetimeplot = False

    fig = plt.figure()

    # The colorbar should only be plotted if more then one event is
    # present.
    if colorbar is not None:
        show_colorbar = colorbar
    else:
        if len(lons) > 1 and hasattr(color, "__len__") and \
                not isinstance(color, (str, native_str)):
            show_colorbar = True
        else:
            show_colorbar = False

    if projection == "local":
        ax_x0, ax_width = 0.10, 0.80
    elif projection == "global":
        ax_x0, ax_width = 0.01, 0.98
    else:
        ax_x0, ax_width = 0.05, 0.90

    proj_kwargs = proj_kwargs or {}
    if projection == 'global':
        proj_kwargs['central_longitude'] = np.mean(lons)
        proj = ccrs.Mollweide(**proj_kwargs)
    elif projection == 'ortho':
        proj_kwargs['central_latitude'] = np.mean(lats)
        proj_kwargs['central_longitude'] = mean_longitude(lons)
        proj = ccrs.Orthographic(**proj_kwargs)
    elif projection == 'local':
        if min(lons) < -150 and max(lons) > 150:
            max_lons = max(np.array(lons) % 360)
            min_lons = min(np.array(lons) % 360)
        else:
            max_lons = max(lons)
            min_lons = min(lons)
        lat_0 = max(lats) / 2. + min(lats) / 2.
        lon_0 = max_lons / 2. + min_lons / 2.
        if lon_0 > 180:
            lon_0 -= 360
        deg2m_lat = 2 * np.pi * 6371 * 1000 / 360
        deg2m_lon = deg2m_lat * np.cos(lat_0 / 180 * np.pi)
        if len(lats) > 1:
            height = (max(lats) - min(lats)) * deg2m_lat
            width = (max_lons - min_lons) * deg2m_lon
            margin = 0.2 * (width + height)
            height += margin
            width += margin
        else:
            height = 2.0 * deg2m_lat
            width = 5.0 * deg2m_lon
        # Do intelligent aspect calculation for local projection
        # adjust to figure dimensions
        w, h = fig.get_size_inches()
        aspect = w / h
        if show_colorbar:
            aspect *= 1.2
        if width / height < aspect:
            width = height * aspect
        else:
            height = width / aspect

        proj_kwargs['central_latitude'] = lat_0
        proj_kwargs['central_longitude'] = lon_0
        proj_kwargs['standard_parallels'] = [lat_0, lat_0]
        proj = ccrs.AlbersEqualArea(**proj_kwargs)

    # User-supplied projection.
    elif isinstance(projection, type):
        if 'central_longitude' in proj_kwargs:
            if proj_kwargs['central_longitude'] == 'auto':
                proj_kwargs['central_longitude'] = mean_longitude(lons)
        if 'central_latitude' in proj_kwargs:
            if proj_kwargs['central_latitude'] == 'auto':
                proj_kwargs['central_latitude'] = np.mean(lats)
        if 'pole_longitude' in proj_kwargs:
            if proj_kwargs['pole_longitude'] == 'auto':
                proj_kwargs['pole_longitude'] = np.mean(lons)
        if 'pole_latitude' in proj_kwargs:
            if proj_kwargs['pole_latitude'] == 'auto':
                proj_kwargs['pole_latitude'] = np.mean(lats)

        proj = projection(**proj_kwargs)

    else:
        msg = "Projection '%s' not supported." % projection
        raise ValueError(msg)

    if show_colorbar:
        map_ax = fig.add_axes([ax_x0, 0.13, ax_width, 0.77], projection=proj)
        cm_ax = fig.add_axes([ax_x0, 0.05, ax_width, 0.05])
        plt.sca(map_ax)
    else:
        ax_y0, ax_height = 0.05, 0.85
        if projection == "local":
            ax_y0 += 0.05
            ax_height -= 0.05
        map_ax = fig.add_axes([ax_x0, ax_y0, ax_width, ax_height],
                              projection=proj)

    if projection == 'local':
        x0, y0 = proj.transform_point(lon_0, lat_0, proj.as_geodetic())
        map_ax.set_xlim(x0 - width / 2, x0 + width / 2)
        map_ax.set_ylim(y0 - height / 2, y0 + height / 2)
    else:
        map_ax.set_global()

    # Pick features at specified resolution.
    resolution = _CARTOPY_RESOLUTIONS[resolution]
    try:
        borders, land, ocean = _CARTOPY_FEATURES[resolution]
    except KeyError:
        borders = cfeature.NaturalEarthFeature(cfeature.BORDERS.category,
                                               cfeature.BORDERS.name,
                                               resolution,
                                               edgecolor='none',
                                               facecolor='none')
        land = cfeature.NaturalEarthFeature(cfeature.LAND.category,
                                            cfeature.LAND.name, resolution,
                                            edgecolor='face', facecolor='none')
        ocean = cfeature.NaturalEarthFeature(cfeature.OCEAN.category,
                                             cfeature.OCEAN.name, resolution,
                                             edgecolor='face',
                                             facecolor='none')
        _CARTOPY_FEATURES[resolution] = (borders, land, ocean)

    # Draw coast lines, country boundaries, fill continents.
    if MATPLOTLIB_VERSION >= [2, 0, 0]:
        map_ax.set_facecolor(water_fill_color)
    else:
        map_ax.set_axis_bgcolor(water_fill_color)
    map_ax.add_feature(ocean, facecolor=water_fill_color)
    map_ax.add_feature(land, facecolor=continent_fill_color)
    map_ax.add_feature(borders, edgecolor='0.75')
    map_ax.coastlines(resolution=resolution, color='0.4')

    # Draw grid lines - TODO: draw_labels=True doesn't work yet.
    if projection == 'local':
        map_ax.gridlines()
    else:
        # Draw lat/lon grid lines every 30 degrees.
        map_ax.gridlines(xlocs=range(-180, 181, 30), ylocs=range(-90, 91, 30))

    # Plot labels
    if labels and len(lons) > 0:
        with map_ax.hold_limits():
            for name, xpt, ypt, _colorpt in zip(labels, lons, lats, color):
                map_ax.text(xpt, ypt, name, weight="heavy", color="k",
                            zorder=100, transform=ccrs.Geodetic(),
                            path_effects=[
                                patheffects.withStroke(linewidth=3,
                                                       foreground="white")])

    scatter = map_ax.scatter(lons, lats, marker=marker, s=size, c=color,
                             zorder=10, cmap=colormap,
                             transform=ccrs.Geodetic())

    if title:
        plt.suptitle(title)

    # Only show the colorbar for more than one event.
    if show_colorbar:
        if colorbar_ticklabel_format is not None:
            if isinstance(colorbar_ticklabel_format, (str, native_str)):
                formatter = FormatStrFormatter(colorbar_ticklabel_format)
            elif hasattr(colorbar_ticklabel_format, '__call__'):
                formatter = FuncFormatter(colorbar_ticklabel_format)
            elif isinstance(colorbar_ticklabel_format, Formatter):
                formatter = colorbar_ticklabel_format
            locator = MaxNLocator(5)
        else:
            if datetimeplot:
                locator = AutoDateLocator()
                formatter = AutoDateFormatter(locator)
                # Compat with old matplotlib versions.
                if hasattr(formatter, "scaled"):
                    formatter.scaled[1 / (24. * 60.)] = '%H:%M:%S'
            else:
                locator = None
                formatter = None
        cb = Colorbar(cm_ax, scatter, cmap=colormap,
                      orientation='horizontal',
                      ticks=locator,
                      format=formatter)
        # Compat with old matplotlib versions.
        if hasattr(cb, "update_ticks"):
            cb.update_ticks()

    if show:
        plt.show()

    return fig


def plot_map(method, *args, **kwargs):
    '''
    Creates a map plot with a data point scatter plot.

    :type method: str
    :param method: Method to use for plotting. Possible values are:

        * ``'basemap'`` to use the Basemap library. For other arguments, see
          the :func:`plot_basemap` function.
        * ``'cartopy'`` to use the Cartopy library. For other arguments, see
          the :func:`plot_cartopy` function.
        * ``None`` to use either the Basemap or Cartopy library, whichever is
          available.
    '''

    if method is None:
        if HAS_BASEMAP:
            return plot_basemap(*args, **kwargs)
        elif HAS_CARTOPY:
            return plot_cartopy(*args, **kwargs)
        else:
            raise ImportError('Neither Basemap nor Cartopy could be imported.')
    elif method == 'basemap':
        if not HAS_BASEMAP:
            raise ImportError('Basemap cannot be imported but was explicitly '
                              'requested.')
        return plot_basemap(*args, **kwargs)
    elif method == 'cartopy':
        if not HAS_CARTOPY:
            raise ImportError('Cartopy cannot be imported but was explicitly '
                              'requested.')
        return plot_cartopy(*args, **kwargs)
    else:
        raise ValueError("The method argument must be either 'basemap' or "
                         "'cartopy', not '%s'." % (method, ))