''' Models for mapping values from one range or space to another.
'''
from __future__ import absolute_import
import warnings
from .. import palettes
from ..core.has_props import abstract
from ..core.properties import Color, Either, Enum, Float, Int, Seq, String, Tuple
from ..core.enums import Palette
from .transforms import Transform
@abstract
[docs]class ColorMapper(Transform):
''' Base class for color mapper types.
'''
palette = Seq(Color, help="""
A sequence of colors to use as the target palette for mapping.
This property can also be set as a ``String``, to the name of any of the
palettes shown in :ref:`bokeh.palettes`.
""").accepts(Enum(Palette), lambda pal: getattr(palettes, pal))
nan_color = Color(default="gray", help="""
Color to be used if data is NaN. Default: 'gray'
""")
def __init__(self, palette=None, **kwargs):
if palette is not None:
kwargs['palette'] = palette
super(ColorMapper, self).__init__(**kwargs)
[docs]class CategoricalColorMapper(ColorMapper):
''' Map categories to colors. Values that are passed to
this mapper that aren't in factors will be assigned the nan_color.
'''
factors = Either(Seq(String), Seq(Tuple(String, String)), Seq(Tuple(String, String, String)), default=None, help="""
A sequence of factors / categories that map to the color palette. For
example the following color mapper:
.. code-block:: python
mapper = CategoricalColorMapper(palette=["red", "blue"], factors=["foo", "bar"])
will map the factor ``"foo"`` to red and the factor ``"bar"`` to blue.
""")
start = Int(default=0, help="""
A start index to "slice" data factors with before color mapping.
For example, if the data to color map consists of 2-level factors such
as ``["2016", "sales"]`` and ``["2016", "marketing"]``, then setting
``start=1`` will perform color mapping only based on the second sub-factor
(i.e. in this case based on the department ``"sales"`` or ``"marketing"``)
""")
end = Int(help="""
A start index to "slice" data factors with before color mapping.
For example, if the data to color map consists of 2-level factors such
as ``["2016", "sales"]`` and ``["2017", "marketing"]``, then setting
``end=1`` will perform color mapping only based on the first sub-factor
(i.e. in this case based on the year ``"2016"`` or ``"2017"``)
If ``None`` then all sub-factors from ``start`` to the end of the
factor will be used for color mapping.
""")
def __init__(self, **kwargs):
super(ColorMapper, self).__init__(**kwargs)
palette = self.palette
factors = self.factors
if palette is not None and factors is not None:
if len(palette) < len(factors):
extra_factors = factors[len(palette):]
warnings.warn("Palette length does not match number of factors. %s will be assigned to `nan_color` %s" % (extra_factors, self.nan_color))
@abstract
[docs]class ContinuousColorMapper(ColorMapper):
''' Base class for continuous color mapper types.
'''
low = Float(help="""
The minimum value of the range to map into the palette. Values below
this are clamped to ``low``.
""")
high = Float(help="""
The maximum value of the range to map into the palette. Values above
this are clamped to ``high``.
""")
low_color = Color(default=None, help="""
Color to be used if data is lower than ``low`` value. If None,
values lower than ``low`` are mapped to the first color in the palette.
""")
high_color = Color(default=None, help="""
Color to be used if data is higher than ``high`` value. If None,
values higher than ``high`` are mapped to the last color in the palette.
""")
[docs]class LinearColorMapper(ContinuousColorMapper):
''' Map numbers in a range [*low*, *high*] linearly into a sequence of
colors (a palette).
For example, if the range is [0, 99] and the palette is
``['red', 'green', 'blue']``, the values would be mapped as follows::
x < 0 : 'red' # values < low are clamped
0 >= x < 33 : 'red'
33 >= x < 66 : 'green'
66 >= x < 99 : 'blue'
99 >= x : 'blue' # values > high are clamped
'''
[docs]class LogColorMapper(ContinuousColorMapper):
''' Map numbers in a range [*low*, *high*] into a sequence of colors
(a palette) on a natural logarithm scale.
For example, if the range is [0, 25] and the palette is
``['red', 'green', 'blue']``, the values would be mapped as follows::
x < 0 : 'red' # values < low are clamped
0 >= x < 2.72 : 'red' # math.e ** 1
2.72 >= x < 7.39 : 'green' # math.e ** 2
7.39 >= x < 20.09 : 'blue' # math.e ** 3
20.09 >= x : 'blue' # values > high are clamped
.. warning::
The LogColorMapper only works for images with scalar values that are
non-negative.
'''
