# Copyright Iris contributors
#
# This file is part of Iris and is released under the LGPL license.
# See COPYING and COPYING.LESSER in the root of the repository for full
# licensing details.
"""
Load, configure and register color map palettes and initialise
color map meta-data mappings.
"""
from functools import wraps
import os
import os.path
import re
import cf_units
from matplotlib import colormaps as mpl_colormaps
import matplotlib.cm as mpl_cm
import matplotlib.colors as mpl_colors
import numpy as np
import iris.config
import iris.cube
# Symmetric normalization function pivot points by SI unit.
PIVOT_BY_UNIT = {cf_units.Unit("K"): 273.15}
# Color map names by palette file metadata field value.
CMAP_BREWER = set()
_CMAP_BY_SCHEME = None
_CMAP_BY_KEYWORD = None
_CMAP_BY_STD_NAME = None
_MISSING_KWARG_CMAP = "missing kwarg cmap"
_MISSING_KWARG_NORM = "missing kwarg norm"
[docs]def is_brewer(cmap):
"""
Determine whether the color map is a Cynthia Brewer color map.
Args:
* cmap:
The color map instance.
Returns:
Boolean.
"""
result = False
if cmap is not None:
result = cmap.name in CMAP_BREWER
return result
def _default_cmap_norm(args, kwargs):
"""
This function injects default cmap and norm behaviour into the keyword
arguments, based on the cube referenced within the positional arguments.
"""
cube = None
# Find the single cube reference within the positional arguments.
for arg in args:
if isinstance(arg, iris.cube.Cube):
cube = arg
break
# Find the keyword arguments of interest.
colors = kwargs.get("colors", None)
# cmap = None to disable default behaviour.
cmap = kwargs.get("cmap", _MISSING_KWARG_CMAP)
# norm = None to disable default behaviour.
norm = kwargs.get("norm", _MISSING_KWARG_NORM)
# Note that "colors" and "cmap" keywords are mutually exclusive.
if colors is None and cube is not None:
std_name = cube.standard_name.lower() if cube.standard_name else ""
# Perform default "cmap" keyword behaviour.
if cmap == _MISSING_KWARG_CMAP:
# Check for an exact match against standard name.
cmaps = _CMAP_BY_STD_NAME.get(std_name, set())
if len(cmaps) == 0:
# Check for a fuzzy match against a keyword.
for keyword in _CMAP_BY_KEYWORD.keys():
if keyword in std_name:
cmaps.update(_CMAP_BY_KEYWORD[keyword])
# Add default color map to keyword arguments.
if len(cmaps):
cmap = sorted(cmaps, reverse=True)[0]
kwargs["cmap"] = mpl_cm.get_cmap(cmap)
# Perform default "norm" keyword behaviour.
if norm == _MISSING_KWARG_NORM:
if "anomaly" in std_name:
# Determine the pivot point.
pivot = PIVOT_BY_UNIT.get(cube.units, 0)
norm = SymmetricNormalize(pivot)
kwargs["norm"] = norm
return args, kwargs
[docs]def cmap_norm(cube):
"""
Determine the default :class:`matplotlib.colors.LinearSegmentedColormap`
and :class:`iris.palette.SymmetricNormalize` instances associated with
the cube.
Args:
* cube (:class:`iris.cube.Cube`):
Source cube to generate default palette from.
Returns:
Tuple of :class:`matplotlib.colors.LinearSegmentedColormap` and
:class:`iris.palette.SymmetricNormalize`
Notes
------
This function maintains laziness when called; it does not realise data.
See more at :doc:`/userguide/real_and_lazy_data`.
"""
args, kwargs = _default_cmap_norm((cube,), {})
return kwargs.get("cmap"), kwargs.get("norm")
[docs]def auto_palette(func):
"""
Decorator wrapper function to control the default behaviour of the
matplotlib cmap and norm keyword arguments.
Args:
* func (callable):
Callable function to be wrapped by the decorator.
Returns:
Closure wrapper function.
"""
@wraps(func)
def wrapper_func(*args, **kwargs):
"""
Closure wrapper function to provide default keyword argument
behaviour.
"""
# Update the keyword arguments with defaults.
args, kwargs = _default_cmap_norm(args, kwargs)
# Call the wrapped function and return its result.
return func(*args, **kwargs)
# Return the closure wrapper function.
return wrapper_func
[docs]class SymmetricNormalize(mpl_colors.Normalize):
"""
Provides a symmetric normalization class around a given pivot point.
"""
def __init__(self, pivot, *args, **kwargs):
self.pivot = pivot
self._vmin = None
self._vmax = None
mpl_colors.Normalize.__init__(self, *args, **kwargs)
def __repr__(self):
return "%s(%r)" % (self.__class__.__name__, self.pivot)
def _update(self, val, update_min=True, update_max=True):
# Update both _vmin and _vmax from given value.
val_diff = np.abs(val - self.pivot)
vmin_diff = np.abs(self._vmin - self.pivot) if self._vmin else 0.0
vmax_diff = np.abs(self._vmax - self.pivot) if self._vmax else 0.0
diff = max(val_diff, vmin_diff, vmax_diff)
if update_min:
self._vmin = self.pivot - diff
if update_max:
self._vmax = self.pivot + diff
@property
def vmin(self):
return getattr(self, "_vmin")
@vmin.setter
def vmin(self, val):
if val is None:
self._vmin = None
elif self._vmax is None:
# Don't set _vmax, it'll stop matplotlib from giving us one.
self._update(val, update_max=False)
else:
# Set both _vmin and _vmax from value
self._update(val)
@property
def vmax(self):
return getattr(self, "_vmax")
@vmax.setter
def vmax(self, val):
if val is None:
self._vmax = None
elif self._vmin is None:
# Don't set _vmin, it'll stop matplotlib from giving us one.
self._update(val, update_min=False)
else:
# Set both _vmin and _vmax from value
self._update(val)
def _load_palette():
"""
Load, configure and register color map palettes and initialise
color map metadata mappings.
"""
# Reference these module level namespace variables.
global CMAP_BREWER, _CMAP_BY_SCHEME, _CMAP_BY_KEYWORD, _CMAP_BY_STD_NAME
_CMAP_BY_SCHEME = {}
_CMAP_BY_KEYWORD = {}
_CMAP_BY_STD_NAME = {}
filenames = []
# Identify all .txt color map palette files.
for root, dirs, files in os.walk(iris.config.PALETTE_PATH):
# Prune any .svn directory from the tree walk.
if ".svn" in dirs:
del dirs[dirs.index(".svn")]
# Identify any target .txt color map palette files.
filenames.extend(
[
os.path.join(root, filename)
for filename in files
if os.path.splitext(filename)[1] == ".txt"
]
)
for filename in filenames:
# Default color map name based on the file base-name (case-SENSITIVE).
cmap_name = os.path.splitext(os.path.basename(filename))[0]
cmap_scheme = None
cmap_keywords = []
cmap_std_names = []
cmap_type = None
# Perform default color map interpolation for quantization
# levels per primary color.
interpolate_flag = True
# Read the file header.
with open(filename) as file_handle:
header = filter(
lambda line: re.match(r"^\s*#.*:\s+.*$", line),
file_handle.readlines(),
)
# Extract the file header metadata.
for line in header:
line = line.replace("#", "", 1).split(":")
head = line[0].strip().lower()
body = line[1].strip()
if head == "name":
# Case-SENSITIVE.
cmap_name = "brewer_{}".format(body)
if head == "scheme":
# Case-insensitive.
cmap_scheme = body.lower()
if head == "keyword":
# Case-insensitive.
keywords = [part.strip().lower() for part in body.split(",")]
cmap_keywords.extend(keywords)
if head == "std_name":
# Case-insensitive.
std_names = [part.strip().lower() for part in body.split(",")]
cmap_std_names.extend(std_names)
if head == "interpolate":
# Case-insensitive.
interpolate_flag = body.lower() != "off"
if head == "type":
# Case-insensitive.
cmap_type = body.lower()
# Integrity check for meta-data 'type' field.
assert cmap_type is not None, (
'Missing meta-data "type" keyword for color map file, "%s"'
% filename
)
assert (
cmap_type == "rgb"
), 'Invalid type [%s] for color map file "%s"' % (cmap_type, filename)
# Update the color map look-up dictionaries.
CMAP_BREWER.add(cmap_name)
if cmap_scheme is not None:
scheme_group = _CMAP_BY_SCHEME.setdefault(cmap_scheme, set())
scheme_group.add(cmap_name)
for keyword in cmap_keywords:
keyword_group = _CMAP_BY_KEYWORD.setdefault(keyword, set())
keyword_group.add(cmap_name)
for std_name in cmap_std_names:
std_name_group = _CMAP_BY_STD_NAME.setdefault(std_name, set())
std_name_group.add(cmap_name)
# Load palette data and create the associated color map.
cmap_data = np.loadtxt(filename)
# Ensure to restrict the number of RGB quantization levels to
# prevent color map interpolation.
if interpolate_flag:
# Perform default color map interpolation for quantization
# levels per primary color.
cmap = mpl_colors.LinearSegmentedColormap.from_list(
cmap_name, cmap_data
)
else:
# Restrict quantization levels per primary color (turn-off
# interpolation).
# Typically used for Brewer color maps.
cmap = mpl_colors.LinearSegmentedColormap.from_list(
cmap_name, cmap_data, N=len(cmap_data)
)
# Register the color map for use.
mpl_colormaps.register(cmap)
# Ensure to load the color map palettes.
_load_palette()
-3.7.0.dev20-gold?style=flat){kind=icon}