def __init__(self, colors, N, stops=None):
# store stops
self.stops = stops
# set colors
self.set_color(colors)
# initialize
Colormap.__init__(self, "test", N)
def __init__(self, colors: Sequence, N: int, stops=None):
""" initialize with the given colors and stops """
# store stops
self.stops = stops
# set colors
self.set_color(colors)
# initialize
Colormap.__init__(self, "test", N)
def plot_energy_error_heatmap(ax,
data_set,
struct_types=struct_types,
dyn_types=dyn_types,
bad_data_traj_list=[],
xbin_size=None,
ybin_size=None,
use_meV_y=False,
use_meV_x=False,
cmap=cm.get_cmap('plasma')):
scalex = 1
scaley = 1
if use_meV_x: scalex = 1000
if use_meV_y: scaley = 1000
def pick_bins(abin_size, A):
if abin_size is None:
abins = np.linspace(A.min(), A.max(), 100)
else:
abins = np.arange(A.min(), A.max(), abin_size)
return abins
#from matplotlib.ticker import MultipleLocator
#ax.yaxis.set_major_locator(MultipleLocator(base=30))
#ax.yaxis.set_minor_locator(MultipleLocator(base=5))
import copy
cmap_tweaked = copy.copy(cmap)
from matplotlib.colors import Colormap
Colormap.set_under(cmap_tweaked, color=(1, 1, 1, 0))
Colormap.set_over(cmap_tweaked, color=(1, 1, 1, 0))
fname = data_set[0]
data_name = data_set[1]
image_pairs = read_evaluation_data(filename=fname,
struct_types=struct_types,
dyn_types=dyn_types,
bad_data_traj_list=bad_data_traj_list)
cache_energy, data_energy = get_energy_lists(image_pairs)
energy_error = cache_energy - data_energy
error_rmse = np.sqrt(np.mean(energy_error**2)) * scaley
error_mae = np.mean(np.absolute(energy_error)) * scaley
X = data_energy * scalex
Y = energy_error * scaley
xbins = pick_bins(xbin_size, X)
ybins = pick_bins(ybin_size, Y)
ax.hist2d(X, Y, bins=(xbins, ybins), vmin=1, cmap=cmap_tweaked)
ax.set_title(data_name, fontsize=8)
ax.minorticks_on()
return error_rmse, error_mae
def plot_force_angle_heatmaps(axes,
data_sets,
struct_types=struct_types,
dyn_types=dyn_types,
struct_colors=struct_colors,
dyn_markers=dyn_markers,
bad_data_traj_list=[],
cmap=cm.get_cmap('plasma')):
deg = 180 / np.pi
bin_size = 2
my_bins = np.arange(0, 180 + bin_size / 2, bin_size)
ylims = (0, 180)
from matplotlib.ticker import MultipleLocator
import copy
cmap_tweaked = copy.copy(cmap)
from matplotlib.colors import Colormap
Colormap.set_under(cmap_tweaked, color=(1, 1, 1, 0))
Colormap.set_over(cmap_tweaked, color=(1, 1, 1, 0))
for di, data_set in enumerate(data_sets):
fname = data_set[0]
data_name = data_set[1]
#color = np.array(get_color(data_set[2]))
#lightness = data_set[3]
#zorder = zorders[di]
image_pairs = read_evaluation_data(filename=fname,
struct_types=struct_types,
dyn_types=dyn_types)
cache_forces, data_forces = get_force_list(image_pairs)
force_cosines_by_atom = compute_force_cosines_by_atom(
cache_forces, data_forces)
data_force_norms_by_atom = compute_force_norms_by_atom(data_forces)
X = collapse_sub_lists(data_force_norms_by_atom)
Y = deg * np.arccos(collapse_sub_lists(force_cosines_by_atom))
xbins = np.linspace(0, X.max(), 100)
axes[di].hist2d(X, Y, bins=(xbins, my_bins), vmin=1, cmap=cmap_tweaked)
#label = data_name + '\nMean: %.2f°\nRMS: %.2f°'%(mean_force_angles, rms_force_angles)
axes[di].set_title(data_name, fontsize=8)
axes[di].set_ylim(ylims)
#ax2.legend(fontsize = 8, handletextpad = 0.3, borderpad = 0.1)
axes[di].minorticks_on()
axes[di].yaxis.set_major_locator(MultipleLocator(base=30))
axes[di].yaxis.set_minor_locator(MultipleLocator(base=5))
if di == 0:
axes[di].set_ylabel('Force Angle, ' + formula_angle + ' (°)')
axes[di].set_xlabel('DFT Force (eV/Å)')
def __init__(self, accels, max_accel=None, min_accel=None):
self.accels = accels
self.points = len(self.accels)
if min_accel is not None:
self.min_accel = min_accel
else:
self.min_accel = np.min(accels)
if max_accel is not None:
self.max_accel = max_accel
else:
self.max_accel = np.max(accels)
Colormap.__init__(self, None, 255)
def __init__(self, source, name='Variable'):
self.name = None # new VariableColormap object
self.bad_set = False
self.set_source(source)
self.monochrome = False
Colormap.__init__(self, name, self.worklut.shape[0]-3)
self.canvases = {}
self.frames = set()
self.slope = 1.0
self.shift = 0.0
self.invrt = 1.0
self.scale = 'LINEAR'
self.auto = True
def __init__(self, cmap):
"""
Make a colormap dynamic with respect to a hinge point.
Dynamic colormaps are stretched to the ``[vmin, vmax]`` range,
by separately scaling the lower part of the colormap
(v < ``hinge``) and the upper part (v > ``hinge``).
For instance, a colormap with v range `[-1, 1]` and hinge at
`0` with `min_color` at `-1`, `hinge_color` and the ``hinge``,
and `max_color` at `1`, can be dynamically scaled to
``[-10, 5]`` with `min_color` at `-10`, `hinge_color` and the
``hinge``, and `max_color` at `5`::
<|min_color-----------hinge_color-----------max_color|>
-1 0 1
\
\
\
\
<|min_color-----------------hinge_color-----max_color|>
-10 0 5
See Also
--------
DynamicColormap.set_range
Scale the colormap to a new range, keeping, or overriding
the hinge point.
"""
self.monochrome = False
Colormap.__init__(self, cmap.name, cmap.N)
try:
self.values = cmap.values
except AttributeError:
self.values = np.linspace(-1, 1, self.N)
self._vmin = self.values[0]
try:
self._hinge = cmap.hinge
except AttributeError:
self._hinge = 0
self._vmax = self.values[-1]
try:
self.colors = cmap.colors
except AttributeError:
self.colors = cmap(self.values)
self.set_range(self.vmin, self.vmax, self.hinge)
self._lut = cmap._lut
self._isinit = True
self._set_extremes()
def __init__(self, source, name='Variable'):
self.name = None # new VariableColormap object
self.bad_set = False
self.set_source(source)
self.monochrome = False
Colormap.__init__(self, name, self.worklut.shape[0]-3)
self.canvases = {}
self.frames = set()
self.slope = 1.0
self.shift = 0.0
self.invrt = 1.0
self.scale = 'LINEAR'
self.auto = True
self.callback = None
def _matplotlib_color(colormap: Colormap, data: float) -> str:
(r, g, b, a) = colormap.__call__(data)
r = round(255 * r)
g = round(255 * g)
b = round(255 * b)
return _rgb_color((r, g, b))
def test_wcs_having_no_naxis(self):
w2 = astropy_wcs.WCS(header={
'WCSAXES': 2, # Number of coordinate axes
'CRPIX1': 1000.0, # Pixel coordinate of reference point
'CRPIX2': 500.0, # Pixel coordinate of reference point
'CDELT1': -0.18, # [deg] Coordinate increment at reference point
'CDELT2': 0.18, # [deg] Coordinate increment at reference point
'CUNIT1': 'deg', # Units of coordinate increment and value
'CUNIT2': 'deg', # Units of coordinate increment and value
'CTYPE1': 'GLON-MOL', # galactic longitude, Mollweide's projection
'CTYPE2': 'GLAT-MOL', # galactic latitude, Mollweide's projection
'CRVAL1': 0.0, # [deg] Coordinate value at reference point
'CRVAL2': 0.0, # [deg] Coordinate value at reference point
})
with pytest.raises(AttributeError):
result = hips2fits.query_with_wcs(
hips=self.hips,
wcs=w2,
get_query_payload=False,
format='jpg',
min_cut=0.5,
max_cut=99.5,
cmap=Colormap('viridis'),
)
def test_query_jpg_no_wcs(self):
from astropy.coordinates import Longitude
from astropy.coordinates import Latitude
from astropy.coordinates import Angle
import astropy.units as u
result = hips2fits.query(
hips=self.hips,
width=1000,
height=500,
fov=Angle(20 * u.deg),
ra=Longitude(0 * u.deg),
dec=Latitude(20 * u.deg),
projection="TAN",
get_query_payload=False,
format='jpg',
min_cut=0.5,
max_cut=99.5,
cmap=Colormap('viridis'),
)
# import matplotlib.cm as cm
# import matplotlib.pyplot as plt
# im = plt.imshow(result)
# plt.show(im)
# We must get a numpy array with 3 dimensions, and the last one should be of size 3 (RGB)
assert isinstance(result, np.ndarray) and result.shape[2] == 3
def __call__(self, value, *args, **kwargs):
# get the color
result = Colormap.__call__(self, value, *args, **kwargs)
# add meta values to it
result = CmapColor(result)
result.setMeta(value, self)
# return the color
return result
def __call__(self, value: float, *args, **kwargs):
""" get the color associated with the given value from the colormap """
# get the color
result = Colormap.__call__(self, value, *args, **kwargs)
# add meta values to it
result = CmapColor(result)
result.setMeta(value, self)
# return the color
return result
def test_bad_strech(self):
with pytest.raises(AttributeError):
result = hips2fits.query_with_wcs(
hips=self.hips,
wcs=self.w,
get_query_payload=False,
format='jpg',
stretch="azs",
min_cut=0.5,
max_cut=99.5,
cmap=Colormap('viridis'),
)
def test_query_jpg(self):
result = hips2fits.query_with_wcs(
hips=self.hips,
wcs=self.w,
get_query_payload=True,
format='jpg',
min_cut=0.5,
max_cut=99.5,
cmap=Colormap('viridis'),
)
# We must get a numpy array with 3 dimensions, and the last one should be of size 3 (RGB)
assert result["format"] == 'jpg' and result["hips"] == "CDS/P/DSS2/red"
def set_colormap(self, colormap, bad_data='grey', update=True):
"""Sets a colormap.
Args:
colormap (str): A color scheme for plotting.
bad_data (str): The color to use for bad data values. Needs to
be a matplotlib color name.
update (bool): Set to True to update the plot.
"""
if isinstance(colormap, str):
colormap = Colormap(colormap)
self.cmap = colormap
if bad_data:
self.cmap.set_bad(color=bad_data)
if update:
self.update()
def test_query_no_wcs_fits(self):
result = hips2fits.query(
hips=self.hips,
get_query_payload=True,
width=1000,
height=500,
fov=Angle(20 * u.deg),
ra=Longitude(0 * u.deg),
dec=Latitude(20 * u.deg),
projection="TAN",
min_cut=0.5,
max_cut=99.5,
cmap=Colormap('viridis'),
)
# We must get a numpy array with 3 dimensions, and the last one should be of size 3 (RGB)
assert result["format"] == 'fits' and result["hips"] == "CDS/P/DSS2/red"
def test_query_jpg(self):
result = hips2fits.query_with_wcs(
hips=self.hips,
wcs=self.w,
get_query_payload=False,
format='jpg',
min_cut=0.5,
max_cut=99.5,
cmap=Colormap('viridis'),
)
# import matplotlib.cm as cm
# import matplotlib.pyplot as plt
# im = plt.imshow(result)
# plt.show(im)
# We must get a numpy array with 3 dimensions, and the last one should be of size 3 (RGB)
assert isinstance(result, np.ndarray) and result.shape[2] == 3
landNAN[:] = np.nan
land = np.hstack((landXY,landNAN))
# Data we want to read and interpolate
data = glob.glob(path+'*.npy')
data.sort()
print data
# Colormap Chl de ref
# COULEUR
norm_chl=mpl.colors.LogNorm(vmin=0.01, vmax=20)
colors = [(0.33,0.33,0.33)] + [(plt.cm.jet(i)) for i in xrange(1,256)]
new_map_chl = mpl.colors.LinearSegmentedColormap.from_list('new_map_chl', colors, N=256)
new_map_chl._init(); new_map_chl._lut[0,:] = new_map_chl._lut[1,:] # Replace lowest value of colormap (which is gray) with the one before (dark blue)
Colormap.set_under(new_map_chl,color=new_map_chl._lut[0,:]) # Set color for values outside colormap to be the lowest of the colormap (dark blue)
##Colormap.set_over(new_map_chl,color=(0.0, 0.0, 0.517825311942959, 1.0))
## to get rgba from colormap for a specific value : new_map_chl(specificvalue for example ex : 0.2)
# BLACK AND WHITE
grays = [(0.33,0.33,0.33)] + [(plt.cm.gray(i)) for i in xrange(1,256)]
new_map_gray_chl = mpl.colors.LinearSegmentedColormap.from_list('new_map_gray_chl', grays, N=256)
# Colormap (avec borne grise)
#norm_chl=mpl.colors.LogNorm(vmin=0.01, vmax=20)
#colors = [(0.33,0.33,0.33)] + [(plt.cm.jet(i)) for i in xrange(1,256)]
#new_map_chl = mpl.colors.LinearSegmentedColormap.from_list('new_map_chl', colors, N=256)
#colors = [(0.33,0.33,0.33)] + [(plt.cm.gray(i)) for i in xrange(1,256)]
#new_map_gray_chl = mpl.colors.LinearSegmentedColormap.from_list('new_map_gray_chl', colors, N=256)
#==============================================================================
# |_____/|_|_| |_| |_| |_| \_\___|\__,_|\__,_|\___|\__|_|\___/|_| |_|
#
################### Reduce dimensions ########################
off_f_red = LLE(n_neighbors=50,
n_components=3).fit_transform(np.transpose(all_off_f))
#off_f_red = LLE(n_neighbors = 50,n_components=3).fit_transform(np.transpose(off_firing[0]))
#off_f_red = TSNE(n_components=3).fit_transform(np.transpose(all_off_f))
## 3D Plot for single trajectory
for i in range(4):
fig = plt.figure()
ax = Axes3D(fig)
trial_len = int((tot_time - window_size) / step_size) - 1
ran_inds = np.arange((trial_len * i), (trial_len * (i + 1)))
this_cmap = Colormap('hsv')
p = ax.scatter(off_f_red[ran_inds, 0],
off_f_red[ran_inds, 1],
off_f_red[ran_inds, 2],
c=np.linspace(1, 255, len(ran_inds)),
cmap='hsv')
ax.plot(off_f_red[ran_inds, 0], off_f_red[ran_inds, 1], off_f_red[ran_inds,
2])
fig.colorbar(p)
## 2D animated scatter plot
###########################
fig, ax = plt.subplots()
x, y = off_f_red[:, 0], off_f_red[:, 1]
sc = ax.scatter([], [])
plt.xlim(min(x), max(x))
def plot_force_error_heatmap(ax,
data_set,
struct_types=struct_types,
dyn_types=dyn_types,
bad_data_traj_list=[],
xbin_size=None,
ybin_size=None,
use_meV_y=False,
use_meV_x=False,
by_atom=False,
cmap=cm.get_cmap('viridis')):
scalex = 1
scaley = 1
if use_meV_x: scalex = 1000
if use_meV_y: scaley = 1000
def pick_bins(abin_size, A):
if abin_size is None:
abins = np.linspace(A.min(), A.max(), 100)
else:
abins = np.arange(A.min(), A.max(), abin_size)
return abins
#from matplotlib.ticker import MultipleLocator
#ax.yaxis.set_major_locator(MultipleLocator(base=30))
#ax.yaxis.set_minor_locator(MultipleLocator(base=5))
import copy
cmap_tweaked = copy.copy(cmap)
from matplotlib.colors import Colormap
Colormap.set_under(cmap_tweaked, color=(1, 1, 1, 0))
Colormap.set_over(cmap_tweaked, color=(1, 1, 1, 0))
fname = data_set[0]
data_name = data_set[1]
image_pairs = read_evaluation_data(filename=fname,
struct_types=struct_types,
dyn_types=dyn_types)
cache_forces, data_forces = get_force_list(image_pairs)
force_error_list = compute_force_error_list(cache_forces, data_forces)
if by_atom:
atom_force_norms = compute_atom_force_norms(data_forces)
rms_force_error_by_atom = compute_rms_force_error_by_atom(
cache_forces, data_forces)
net_rms_force_error_by_atom = np.sqrt(
np.mean(collapse_sub_lists(rms_force_error_by_atom)**2))
X = scalex * atom_force_norms
Y = scaley * rms_force_error_by_atom
error_rmse = scaley * net_rms_force_error_by_atom
else:
image_force_norms = compute_force_norms_by_image(
data_forces) / np.sqrt(3)
rms_force_error_by_image = compute_rms_force_error_by_image(
cache_forces, data_forces) / np.sqrt(3)
net_rms_force_error_by_image = np.sqrt(
np.mean((rms_force_error_by_image)**2))
X = scalex * image_force_norms
Y = scaley * rms_force_error_by_image
error_rmse = scaley * net_rms_force_error_by_image
xbins = pick_bins(xbin_size, X)
ybins = pick_bins(ybin_size, Y)
ax.hist2d(X, Y, bins=(xbins, ybins), vmin=1, cmap=cmap_tweaked)
ax.set_title(data_name, fontsize=8)
ax.minorticks_on()
return error_rmse
def __init__(self, name, color, min_alpha = 0.0, max_alpha = 1.0, N=256):
Colormap.__init__(self, name, N)
self._color = color
self._min_alpha = min_alpha
self._max_alpha = max_alpha
}
return pd.DataFrame(d)
init_data()
raw = row_data()
cols = 3
rows = len(data) / cols + (1 if len(data) % cols != 0 else 0)
colorFrom = ''
colorTo = ''
# with open('color.txt', 'r') as fin:
# colorFrom, colorTo = fin.readline().split(' ')
cmap = Colormap('current')
# cmap.set_over(color=colorFrom)
# cmap.set_under(color=colorTo)
figure = plt.figure()
def xy_by_plot(x, y):
col = int(x + 0.5)
row = int(y + 0.5)
return col, row
for i, key in enumerate(data):
vals = data[key]
def plot_map(self, argv):
self.geo = Geometry(model='load')
(Lx, Ly) = self.geo.get_repeated_size(argv)
Sx = 8
Sy = Sx * Ly / Lx
fig = figure(num=' ',
figsize=(Sx, Sy),
dpi=80,
facecolor='w',
edgecolor='k')
axes([0, 0, 1.0, 1.0])
#fig = figure(num=' ', figsize=(8*Lx/Ly, 8), dpi=80, facecolor='w', edgecolor='k')
#axes([0,0,0.5,1.0])
#axes([0,0,1.0,1.0])
#data += 1.0
plt.set_cmap(Colormap('hot'))
(data, nodes, triangulation) = self.get_data(argv)
vmin = argv.setdefault('vmin', min(data))
vmax = argv.setdefault('vmax', max(data))
tripcolor(triangulation,
np.array(data),
shading='gouraud',
norm=mpl.colors.Normalize(vmin=vmin, vmax=vmax),
zorder=1)
#colorbar(norm=mpl.colors.Normalize(vmin=min(data),vmax=max(data)))
#Contour-----
if argv.setdefault('iso_values', False):
t = tricontour(triangulation,
data,
levels=np.linspace(vmin, vmax, 10),
colors='black',
linewidths=1.5)
if argv.setdefault(
'streamlines', False
): # and (variable == 'fourier_flux' or variable == 'flux' ):
n_lines = argv.setdefault(
'n_lines', 10) * self.Ny #assuming transport across x
xi = np.linspace(-Lx * 0.5, Lx * 0.5, 300 * self.Nx)
yi = np.linspace(-Ly * 0.5, Ly * 0.5, 300 * self.Ny)
x = np.array(nodes)[:, 0]
y = np.array(nodes)[:, 1]
z = np.array(data).T[0]
Fx = griddata((x, y),
z, (xi[None, :], yi[:, None]),
method='cubic')
z = np.array(data).T[1]
Fy = griddata((x, y),
z, (xi[None, :], yi[:, None]),
method='cubic')
seed_points = np.array([
-Lx * 0.5 * 0.99 * np.ones(n_lines),
np.linspace(-Ly * 0.5 * 0.98, Ly * 0.5 * 0.98, n_lines)
])
ss = streamplot(xi,
yi,
Fx,
Fy,
maxlength=1e8,
start_points=seed_points.T,
integration_direction='both',
color='r',
minlength=0.95,
linewidth=1)
if argv.setdefault('plot_interfaces', False):
pp = self.geo.get_interface_point_couples(argv)
axis('off')
gca().invert_yaxis()
xlim([-Lx * 0.5, Lx * 0.5])
ylim([-Ly * 0.5, Ly * 0.5])
gca().margins(x=0, y=0)
axis('tight')
axis('equal')
savefig('fig.png', dpi=200)
show()
def plot_force_angle_polar_heatmap(
ax,
data_set,
struct_types=struct_types,
dyn_types=dyn_types,
bad_data_traj_list=[],
theta_zero_location="S",
xbin_size=None,
ybin_size=None,
use_meV_y=False,
#use_meV_x = False,
by_atom=True,
cmap=cm.get_cmap('plasma')):
scalex = 1
scaley = 1
#if use_meV_x: scalex = 1000
if use_meV_y: scaley = 1000
deg = 180 / np.pi
def pick_bins(abin_size, A):
if abin_size is None:
abins = np.linspace(np.min(A), np.max(A), 100)
else:
abins = np.arange(np.min(A), np.max(A), abin_size)
return abins
if xbin_size is None:
xbin_size = 2
xbins = np.arange(0, 180, xbin_size)
from matplotlib.ticker import MultipleLocator
#ax.yaxis.set_major_locator(MultipleLocator(base=30))
#ax.yaxis.set_minor_locator(MultipleLocator(base=5))
import copy
cmap_tweaked = copy.copy(cmap)
from matplotlib.colors import Colormap
Colormap.set_under(cmap_tweaked, color=(1, 1, 1, 0))
Colormap.set_over(cmap_tweaked, color=(1, 1, 1, 0))
fname = data_set[0]
data_name = data_set[1]
image_pairs = read_evaluation_data(filename=fname,
struct_types=struct_types,
dyn_types=dyn_types)
cache_forces, data_forces = get_force_list(image_pairs)
force_error_list = compute_force_error_list(cache_forces, data_forces)
if by_atom:
force_norms_by_atom = compute_force_norms_by_atom(data_forces)
rms_force_error_by_atom = compute_rms_force_error_by_atom(
cache_forces, data_forces)
#net_rms_force_error_by_atom = np.sqrt(np.mean( collapse_sub_lists(rms_force_error_by_atom)**2))
force_cosines_by_atom = compute_force_cosines_by_atom(
cache_forces, data_forces)
force_angles_by_atom = deg * np.arccos(
collapse_sub_lists(force_cosines_by_atom))
mean_force_angles = np.mean(force_angles_by_atom)
rms_force_angles = np.sqrt(np.mean((force_angles_by_atom)**2))
X = scalex * force_angles_by_atom
Y = scaley * collapse_sub_lists(rms_force_error_by_atom)
error_rmse = scaley * rms_force_angles
error_mae = scaley * mean_force_angles
else:
image_force_norms = compute_force_norms_by_image(
data_forces) / np.sqrt(3)
rms_force_error_by_image = compute_rms_force_error_by_image(
cache_forces, data_forces) / np.sqrt(3)
net_rms_force_error_by_image = np.sqrt(
np.mean((rms_force_error_by_image)**2))
X = scalex * image_force_norms
Y = scaley * rms_force_error_by_image
error_rmse = scaley * net_rms_force_error_by_image
#xbins = pick_bins(xbin_size, X)
ybins = pick_bins(ybin_size, Y)
ax.hist2d(X / deg, Y, bins=(xbins / deg, ybins), vmin=1, cmap=cmap_tweaked)
ax.set_title(data_name, fontsize=8)
#ax_force_polar.legend(fontsize = 8)
ax.set_thetamin(0)
ax.set_thetamax(180)
ax.set_theta_zero_location(theta_zero_location)
ax.xaxis.set_major_locator(MultipleLocator(base=45 / deg))
#ax.xaxis.set_minor_locator(MultipleLocator(base=15/deg))
dtheta = 15
theta_grid = np.arange(0, 180 + dtheta / 2, dtheta)
for theta in theta_grid:
ax.axvline(theta / deg, color='grey', lw=0.8, zorder=-1)
#ax.set_thetagrids()
#ax.minorticks_on()
return error_rmse, error_mae
def __call__(self, X, alpha=1.0, bytes=False):
if self.bad_set:
if not isinstance(X, numpy.ma.masked_array):
X = numpy.ma.asarray(X)
X.mask = ma.make_mask(~numpy.isfinite(X))
return Colormap.__call__(self, X, alpha, bytes)
def __init__(self, name, color, min_alpha=0.0, max_alpha=1.0, N=256):
Colormap.__init__(self, name, N)
self._color = color
self._min_alpha = min_alpha
self._max_alpha = max_alpha
def __init__(self, cmap, vmin=0, vmax=1):
if not has_mpl:
raise ImportError("Could not import all matplotlib dependencies!")
if isinstance(cmap, str):
Colormap.__init__(self, cmap, vmin, vmax)
cmap = mpl_get_cmap(cmap)
else:
try:
name = str(cmap.name)
except:
raise ValueError("The argument 'cmap' is of wrong type!")
Colormap.__init__(self, name, vmin, vmax)
# Obtain table to convert to .cpt:
if 'colors' in cmap.__dict__.keys():
if isinstance(cmap.colors, list):
N = len(cmap.colors)
elif isinstance(cmap.colors, np.ndarray):
N = cmap.colors.shape[0]
else:
raise ValueError("Data type not understood: " +
str(type(cmap.colors)))
self.cpt_table = np.zeros((N, 4))
self.cpt_table[:, 1:4] = cmap.colors
self.cpt_table[:, 0] = np.linspace(0, 1, N)
elif '_segmentdata' in cmap.__dict__.keys():
# Obtain the segments for all three colors:
segs = [
np.array(cmap._segmentdata['red']),
np.array(cmap._segmentdata['green']),
np.array(cmap._segmentdata['blue'])
]
changes = set([seg[0] for seg in segs[0]]) \
| set([seg[0] for seg in segs[1]]) \
| set([seg[0] for seg in segs[2]])
changes = np.sort(np.array(list(changes)))[1:]
colors = []
colors += [[0, segs[0][0, 1], segs[1][0, 1], segs[2][0, 1]]]
for x in changes:
c0 = [0, 0, 0]
c1 = [0, 0, 0]
for k in range(3):
id_ = np.argwhere(segs[k][:, 0] < x)[-1]
if id_ < len(segs[k]):
if segs[k][id_ + 1, 0] == x:
# Red value changes at x:
c0[k] = segs[k][id_ + 1, 1][0]
c1[k] = segs[k][id_ + 1, 2][0]
else:
# Red value does not change at x. Interpolate linearly:
x0 = segs[k][id_, 0][0]
v0 = segs[k][id_, 2][0]
x1 = segs[k][id_ + 1, 0][0]
v1 = segs[k][id_ + 1, 1][0]
c0[k] = (x - x0) / (x1 - x0) * v0 + (x1 - x) / (
x1 - x0) * v1
c1[k] = c0[k]
colors += [[x, c0[0], c0[1], c0[2]]]
if c0[0] != c1[0] or c0[0] != c1[0] or c0[0] != c1[0]:
colors += [[x, c1[0], c1[1], c1[2]]]
self.cpt_table = np.array(colors)
def plot_choropleth(fig,
axg,
gdf,
column,
cmap=cmap_turbo,
vmin=None,
vmax=None,
norm=None,
cticks=None,
discrete=False,
plot_kwargs={},
cbar_kwargs={},
cbar_pos={}):
# colormap
dmin, dmax = np.nanmin(gdf.loc[:, column].values), np.nanmax(
gdf.loc[:, column].values)
vmin = vmin if (vmin is not None) else dmin
vmax = vmax if (vmax is not None) else dmax
if cticks is None:
k = cbar_kwargs.get('k', 10)
cticks = np.linspace(vmin, vmax, k).tolist()
if discrete: #only linear
norm = BoundaryNorm(cticks, cmap.N)
elif norm is None:
norm = Normalize(vmin=vmin, vmax=vmax)
if isinstance(cmap, str):
cmap = Colormap(cmap)
# add choropleth layer to fig
gdf.plot(ax=axg,
column=column,
cmap=cmap,
norm=norm,
vmin=vmin,
vmax=vmax,
**plot_kwargs)
# plot colorbar
cax = None
if cbar_kwargs:
# fig = plt.gcf()
location = cbar_kwargs.pop('location', 'bottom')
clabel = cbar_kwargs.pop('label', '')
extend = cbar_kwargs.pop('extend', None)
pad = cbar_pos.get('pad', 0.05)
fraction = cbar_pos.get('fraction', 0.02)
shrink = cbar_pos.get('shrink', 1)
if extend == 'both' or (extend is None and ((dmax > vmax) and
(dmin < vmin))):
cbar_kwargs.update(extend='both',
ticks=cticks,
boundaries=[dmin] + cticks +
[dmax] if discrete else None)
elif extend == 'max' or (extend is None and dmax > vmax):
cbar_kwargs.update(extend='max',
ticks=cticks,
boundaries=cticks +
[dmax] if discrete else None)
elif extend == 'min' or (extend is None and dmin < vmin):
cbar_kwargs.update(extend='min',
ticks=cticks,
boundaries=[dmin] +
cticks if discrete else None)
else:
cbar_kwargs.update(ticks=cticks,
boundaries=cticks if discrete else None)
gpos = axg.get_position()
if location == 'right':
cax = fig.add_axes([
gpos.x1 + pad, gpos.y0 + gpos.height * (1 - shrink) / 2.,
gpos.width * fraction, gpos.height * shrink
]) # new ax
cbar = mpl.colorbar.Colorbar(
cax, mpl.cm.ScalarMappable(cmap=cmap, norm=norm),
**cbar_kwargs)
cbar.ax.set_ylabel(clabel, rotation='vertical')
elif location == 'bottom':
cax = fig.add_axes([
gpos.x0 + gpos.width * (1 - shrink) / 2., gpos.y0 - pad,
gpos.width * shrink, gpos.height * fraction
]) # new ax
cbar = mpl.colorbar.Colorbar(cax,
mpl.cm.ScalarMappable(cmap=cmap,
norm=norm),
orientation='horizontal',
**cbar_kwargs)
cbar.ax.set_xlabel(clabel)
else:
raise ValueError('unknown value for "location"')
return cax
def set_bad(self, color='k', alpha = 1.0):
self.bad_set = True
self.bad_val = (color, alpha)
Colormap.set_bad(self, color, alpha)
if self.auto:
self.update()
def __init__(self, name, color, N=256):
Colormap.__init__(self, name, N)
self.color = colorConverter.to_rgb(color)