def plot_xy_color_swatch(xy, fig=None, ax=None):
"""Plot a color swatch for a given pair of xy chromaticity coordinates.
Parameters
----------
xy : `numpy.ndarray`
ndarray of x,y chromaticity coordinates
fig : `matplotlib.figure.Figure`
Figure to draw plot in
ax : `matplotlib.axes.Axis`
Axis to draw plot in
Returns
-------
fig : `matplotlib.figure.Figure`
Figure to draw plot in
ax : `matplotlib.axes.Axis`
Axis to draw plot in
"""
xyz = xy_to_XYZ(xy)
rgb = XYZ_to_sRGB(xyz)
imgarr = np.ones(
(2, 2, 3)
) * rgb # * 3.975 # 100 transmission => peak of 0.25, * 4 rescales values to fill [0,1]
imgarr = (imgarr * 3.95 * 255).astype(np.uint8)
fig, ax = share_fig_ax(fig, ax)
ax.imshow(imgarr, interpolation=None)
ax.xaxis.set_major_locator(NullLocator())
ax.yaxis.set_major_locator(NullLocator())
return fig, ax
def plot_cmf(cmf, fig=None, ax=None):
fig, ax = share_fig_ax(fig, ax)
ax.plot(cmf['wvl'], cmf['X'])
ax.plot(cmf['wvl'], cmf['Y'])
ax.plot(cmf['wvl'], cmf['Z'])
return fig, ax
def cie_1976_wavelength_annotations(wavelengths, fig=None, ax=None):
"""Draw lines normal to the spectral locust on a CIE 1976 diagram and writes the text for each wavelength.
Parameters
----------
wavelengths : `iterable`
set of wavelengths to annotate
fig : `matplotlib.figure.Figure`
Figure to draw plot in
ax : `matplotlib.axes.Axis`
Axis to draw plot in
Returns
-------
fig : `matplotlib.figure.Figure`
Figure to draw plot in
ax : `matplotlib.axes.Axis`
Axis to draw plot in
Notes
-----
see SE:
https://stackoverflow.com/questions/26768934/annotation-along-a-curve-in-matplotlib
"""
# some tick parameters
tick_length = 0.025
text_offset = 0.06
# convert wavelength to u' v' coordinates
wavelengths = np.asarray(wavelengths)
idx = np.arange(1, len(wavelengths) - 1, dtype=int)
wvl_lbl = wavelengths[idx]
uv = XYZ_to_uvprime(wavelength_to_XYZ(wavelengths))
u, v = uv[..., 0][idx], uv[..., 1][idx]
u_last, v_last = uv[..., 0][idx - 1], uv[..., 1][idx - 1]
u_next, v_next = uv[..., 0][idx + 1], uv[..., 1][idx + 1]
angle = atan2(v_next - v_last, u_next - u_last) + pi / 2
cos_ang, sin_ang = cos(angle), sin(angle)
u1, v1 = u + tick_length * cos_ang, v + tick_length * sin_ang
u2, v2 = u + text_offset * cos_ang, v + text_offset * sin_ang
fig, ax = share_fig_ax(fig, ax)
tick_lines = LineCollection(np.c_[u, v, u1, v1].reshape(-1, 2, 2),
color='0.25',
lw=1.25)
ax.add_collection(tick_lines)
for i in range(len(idx)):
ax.text(u2[i],
v2[i],
str(wvl_lbl[i]),
va="center",
ha="center",
clip_on=True)
return fig, ax
def plot_spectrum(spectrum_dict,
xrange=(380, 730),
yrange=(0, 100),
smoothing=None,
fig=None,
ax=None):
"""Plot a spectrum.
Parameters
----------
spectrum_dict : `dict`
with keys wvl, values
xrange : `iterable`
pair of lower and upper x bounds
yrange : `iterable`
pair of lower and upper y bounds
smoothing : `float`
number of nanometers to smooth data by. If None, do no smoothing
fig : `matplotlib.figure.Figure`
Figure to draw plot in
ax : `matplotlib.axes.Axis`
Axis to draw plot in
Returns
-------
fig : `matplotlib.figure.Figure`
Figure to draw plot in
ax : `matplotlib.axes.Axis`
Axis to draw plot in
"""
wvl, values = spectrum_dict['wvl'], spectrum_dict['values']
if smoothing is not None:
dx = wvl[1] - wvl[0]
window_width = int(smoothing / dx)
values = smooth(values, window_width, window='flat')
bg = render_plot_spectrum_background(xrange[0], xrange[1])
fig, ax = share_fig_ax(fig, ax)
ax.imshow(bg,
extent=[*xrange, *yrange],
interpolation='lanczos',
aspect='auto')
ax.plot(wvl, values, lw=3)
ax.fill_between(wvl,
values,
yrange[1] * len(values),
facecolor='w',
alpha=0.5)
ax.set(xlim=xrange,
xlabel=r'Wavelength $\lambda$ [nm]',
ylim=yrange,
ylabel='Transmission [%]')
return fig, ax
def cct_duv_diagram(samples=100, fig=None, ax=None):
"""Create a CCT-Duv diagram.
For more information see Calculation of CCT and Duv and Practical Conversion Formulae, Yoshi Ohno, 2011.
Parameters
----------
samples : `int`
number of samples on the background, total #pix will be samples^2
fig : `matplotlib.figure.Figure`
Figure to draw plot in
ax : `matplotlib.axes.Axis`
Axis to draw plot in
Returns
-------
fig : `matplotlib.figure.Figure`
Figure to draw plot in
ax : `matplotlib.axes.Axis`
Axis to draw plot in
"""
xlim = (2000, 10000)
ylim = (-0.03, 0.03)
cct = np.linspace(xlim[0], xlim[1],
samples) # todo: even sampling along log, not linear
duv = np.linspace(ylim[0], ylim[1], samples)
upvp = multi_cct_duv_to_uvprime(cct, duv)
cct, duv = np.meshgrid(cct, duv)
xy = uvprime_to_xy(upvp)
xyz = xy_to_XYZ(xy)
dat = XYZ_to_sRGB(xyz)
maximum = np.max(dat, axis=-1)
dat /= maximum[..., np.newaxis]
dat = np.clip(dat, 0, 1)
fig, ax = share_fig_ax(fig, ax)
ax.imshow(dat,
extent=[*xlim, *ylim],
interpolation='bilinear',
origin='lower',
aspect='auto')
ax.set(xlim=xlim, xlabel='CCT [K]', ylim=ylim, ylabel='Duv [a.u.]')
return fig, ax
def cie_1976_plankian_locust(trange=(2000, 10000),
num_points=100,
isotemperature_lines_at=None,
isotemperature_du=0.025,
fig=None,
ax=None):
"""Draw the plankian locust on the CIE 1976 color diagram.
Parameters
----------
trange : `iterable`
(min,max) color temperatures
num_points : `int`
number of points to compute
isotemperature_lines_at : `iterable`
CCTs to plot isotemperature lines at, defaults to [2000, 3000, 4000, 5000, 6500, 10000] if None.
set to False to not plot lines
isotemperature_du : `float`
delta-u, parameter, length in x of the isotemperature lines
fig : `matplotlib.figure.Figure`
Figure to draw plot in
ax : `matplotlib.axes.Axis`
Axis to draw plot in
Returns
-------
fig : `matplotlib.figure.Figure`
Figure to draw plot in
ax : `matplotlib.axes.Axis`
Axis to draw plot in
"""
# compute the u', v' coordinates of the temperatures
temps = np.linspace(trange[0], trange[1], num_points)
interpf_u, interpf_v = prepare_robertson_interpfs(values=('u', 'v'),
vs='K')
u = interpf_u(temps)
v = interpf_v(temps) * 1.5 # x1.5 converts 1960 uv to 1976 u' v'
# if plotting isotemperature lines, compute the upper and lower points of
# each line and connect them.
plot_isotemp = True
if isotemperature_lines_at is None:
isotemperature_lines_at = np.asarray(
[2000, 3000, 4000, 5000, 6500, 10000])
u_iso = interpf_u(isotemperature_lines_at)
v_iso = interpf_v(isotemperature_lines_at)
interpf_dvdu = prepare_robertson_interpfs(values='dvdu', vs='u')
dvdu = interpf_dvdu(u_iso)
du = isotemperature_du / dvdu
u_high = u_iso + du / 2
u_low = u_iso - du / 2
v_high = (v_iso + du / 2 *
dvdu) * 1.5 # factors of 1.5 convert from uv to u'v'
v_low = (v_iso - du / 2 * dvdu) * 1.5
elif isotemperature_lines_at is False:
plot_isotemp = False
fig, ax = share_fig_ax(fig, ax)
ax.plot(u, v, c='0.15')
if plot_isotemp is True:
for ul, uh, vl, vh in zip(u_low, u_high, v_low, v_high):
ax.plot([ul, uh], [vl, vh], c='0.15')
return fig, ax
def cie_1976_plot(xlim=(-0.09, 0.68),
ylim=None,
samples=400,
annotate_wvl=True,
draw_plankian_locust=False,
fig=None,
ax=None):
"""Create a CIE 1976 plot.
Parameters
----------
xlim : `iterable`
left and right bounds of the plot
ylim : `iterable`
lower and upper bounds of the plot. If `None`, the y bounds will be chosen to match the x bounds
samples : `int`
number of 1D samples within the region of interest, total pixels will be samples^2
annotate_wvl : `bool`
whether to plot wavelength annotations
draw_plankian_locust : `bool`
whether to draw the plankian locust
fig : `matplotlib.figure.Figure`
Figure to draw plot in
ax : `matplotlib.axes.Axis`
Axis to draw plot in
Returns
-------
fig : `matplotlib.figure.Figure`
Figure to draw plot in
ax : `matplotlib.axes.Axis`
Axis to draw plot in
"""
# duplicate xlim if ylim not set
if ylim is None:
ylim = xlim
# don't compute over dead space
xlim_bg = list(xlim)
ylim_bg = list(ylim)
if xlim[0] < 0:
xlim_bg[0] = 0
if xlim[1] > 0.65:
xlim_bg[1] = 0.65
if ylim[0] < 0:
ylim_bg[0] = 0
if ylim[1] > 0.6:
ylim_bg[1] = 0.6
# create lists of wavelengths and map them to uv for the border line and annotation.
wvl_line = np.arange(400, 700, 2)
wvl_line_uv = XYZ_to_uvprime(wavelength_to_XYZ(wvl_line))
# duplicate the lowest wavelength so that the boundary line is closed
wvl_line_uv = np.vstack((wvl_line_uv, wvl_line_uv[0, :]))
background = render_cie_1976_background(*xlim_bg, *ylim_bg, samples)
fig, ax = share_fig_ax(fig, ax)
ax.imshow(background,
extent=[*xlim_bg, *ylim_bg],
interpolation='bilinear',
origin='lower')
ax.plot(wvl_line_uv[:, 0], wvl_line_uv[:, 1], ls='-', c='0.25', lw=2.5)
if annotate_wvl:
wvl_annotate = [
360, 400, 455, 470, 480, 490, 500, 510, 520, 540, 555, 570, 580,
590, 600, 610, 625, 700, 830
]
fig, ax = cie_1976_wavelength_annotations(wvl_annotate, fig=fig, ax=ax)
if draw_plankian_locust:
fig, ax = cie_1976_plankian_locust(fig=fig, ax=ax)
ax.set(xlim=xlim, xlabel='CIE u\'', ylim=ylim, ylabel='CIE v\'')
return fig, ax
def cie_1931_plot(xlim=(0, 0.9), ylim=None, samples=300, fig=None, ax=None):
"""Create a CIE 1931 plot.
Parameters
----------
xlim : `iterable`
left and right bounds of the plot
ylim : `iterable`
lower and upper bounds of the plot. If `None`, the y bounds will be chosen to match the x bounds
samples : `int`
number of 1D samples within the region of interest, total pixels will be samples^2
fig : `matplotlib.figure.Figure`
Figure to draw plot in
ax : `matplotlib.axes.Axis`
Axis to draw plot in
Returns
-------
fig : `matplotlib.figure.Figure`
Figure to draw plot in
ax : `matplotlib.axes.Axis`
Axis to draw plot in
"""
# duplicate xlim if ylim not set
if ylim is None:
ylim = xlim
# don't compute over dead space
xlim_bg = list(xlim)
ylim_bg = list(ylim)
if xlim[0] < 0:
xlim_bg[0] = 0
if xlim[1] > 0.75:
xlim_bg[1] = 0.75
if ylim[0] < 0:
ylim_bg[0] = 0
if ylim[1] > 0.85:
ylim_bg[1] = 0.85
# create lists of wavelengths and map them to uv,
# a reduced set for a faster mask and
# yet another set for annotation.
wvl_line = np.arange(400, 700, 2.5)
wvl_line_xy = XYZ_to_xy(wavelength_to_XYZ(wvl_line))
wvl_annotate = [
360, 400, 455, 470, 480, 490, 500, 510, 520, 540, 555, 570, 580, 590,
600, 615, 630, 700, 830
]
data = render_cie_1931_background(*xlim_bg, *ylim_bg, samples)
# duplicate the lowest wavelength so that the boundary line is closed
wvl_line_xy = np.vstack((wvl_line_xy, wvl_line_xy[0, :]))
fig, ax = share_fig_ax(fig, ax)
ax.imshow(data,
extent=[*xlim_bg, *ylim_bg],
interpolation='bilinear',
origin='lower')
ax.plot(wvl_line_xy[:, 0], wvl_line_xy[:, 1], ls='-', c='0.25', lw=2)
fig, ax = cie_1931_wavelength_annotations(wvl_annotate, fig=fig, ax=ax)
ax.set(xlim=xlim, xlabel='CIE x', ylim=ylim, ylabel='CIE y')
return fig, ax