def add_color(self, color=None):
'''
'''
if color is None:
color = self.labColorRep.cget('bg')
if self.type.get() == 'Qualitative':
if color in list(self.saveColors.values()):
tk.messagebox.showinfo(
'Alread iny ..',
'Color already in ' + 'palette. You can delete colors ' +
'by right-clicking on bars. To change ' +
'the psoition of a color move the bars' +
' to the correct position in the graph.',
parent=self.toplevel)
return
for n in range(len(self.bars) + 5):
if n not in self.saveColors and n in self.bars:
break
elif n not in self.bars:
self.add_rectangle(n)
break
self.bars[n].set_facecolor(color)
self.saveColors[n] = color
self.figure.canvas.draw()
elif 'Gradient' in self.type.get():
if 'sequential' in self.type.get():
self.get_cmap_by_color(color)
self.color_bars_by_gradient()
else:
if self.firstColor is None:
self.firstColor = color
self.firstColorSaved = color
self.get_cmap_by_color(color, reverse=True)
self.color_bars_by_gradient(how='half')
else:
col1 = husl.hex_to_husl(self.firstColor)
col2 = husl.hex_to_husl(color)
s, l = self.get_sl_from_entry()
if s is None:
return
cmapOutput = sns.diverging_palette(
col1[0],
col2[0],
s=s,
l=l,
as_cmap=True,
center=self.center.get())
self.cmap = matplotlib.cm.get_cmap(cmapOutput)
self.firstColor = None
self.color_bars_by_gradient()
def test_snapshot(self):
for hex_color, colors in self.snapshot.items():
# Test forward functions
test_rgb = husl.hex_to_rgb(hex_color)
self.assertTuplesClose(test_rgb, colors['rgb'])
test_xyz = husl.rgb_to_xyz(test_rgb)
self.assertTuplesClose(test_xyz, colors['xyz'])
test_luv = husl.xyz_to_luv(test_xyz)
self.assertTuplesClose(test_luv, colors['luv'])
test_lch = husl.luv_to_lch(test_luv)
self.assertTuplesClose(test_lch, colors['lch'])
test_husl = husl.lch_to_husl(test_lch)
self.assertTuplesClose(test_husl, colors['husl'])
test_huslp = husl.lch_to_huslp(test_lch)
self.assertTuplesClose(test_huslp, colors['huslp'])
# Test backward functions
test_lch = husl.husl_to_lch(colors['husl'])
self.assertTuplesClose(test_lch, colors['lch'])
test_lch = husl.huslp_to_lch(colors['huslp'])
self.assertTuplesClose(test_lch, colors['lch'])
test_luv = husl.lch_to_luv(test_lch)
self.assertTuplesClose(test_luv, colors['luv'])
test_xyz = husl.luv_to_xyz(test_luv)
self.assertTuplesClose(test_xyz, colors['xyz'])
test_rgb = husl.xyz_to_rgb(test_xyz)
self.assertTuplesClose(test_rgb, colors['rgb'])
self.assertEqual(husl.rgb_to_hex(test_rgb), hex_color)
# Full test
self.assertEqual(husl.husl_to_hex(*colors['husl']), hex_color)
self.assertTuplesClose(husl.hex_to_husl(hex_color), colors['husl'])
self.assertEqual(husl.huslp_to_hex(*colors['huslp']), hex_color)
self.assertTuplesClose(husl.hex_to_huslp(hex_color), colors['huslp'])
def update_sl(self, color):
'''
'''
if color is not None:
huslColor = husl.hex_to_husl(color)
self.slVars['s'].set(huslColor[1])
self.slVars['l'].set(huslColor[2])
def test_snapshot(self):
for h in range(37):
for s in range(21):
for l in range(21):
H = h * 10.0
S = s * 5.0
L = l * 5.0
test = husl.husl_to_hex(H, S, L)
correct = self.snapshot['husl'][h][s][l]
self.assertEqual(test, correct)
test2 = husl.husl_to_hex(*husl.hex_to_husl(correct))
self.assertEqual(test2, correct)
def test_snapshot(self):
for h in range(37):
for s in range(21):
for l in range(21):
H = h * 10.0
S = s * 5.0
L = l * 5.0
test = husl.husl_to_hex(H, S, L)
correct = self.snapshot['husl'][h][s][l]
self.assertEqual(test, correct)
test2 = husl.husl_to_hex(*husl.hex_to_husl(correct))
self.assertEqual(test2, correct)
def update_hex_by_husl(self, event):
'''
'''
color = self.labColorRep.cget('bg')
huslColorHue = husl.hex_to_husl(color)[0]
s, l = self.get_sl_from_entry()
if s is None:
return
hex = husl.husl_to_hex(huslColorHue, s, l)
self.labColorRep.configure(bg=hex)
paletteType = self.type.get()
if paletteType != 'Qualitative':
if 'diverging' in paletteType:
self.firstColor = str(self.firstColorSaved)
self.add_color()
else:
self.add_color()
def test_snapshot(self):
for hex_color, colors in self.snapshot.items():
# Test forward functions
test_rgb = husl.hex_to_rgb(hex_color)
self.assertTuplesClose(test_rgb, colors['rgb'])
test_xyz = husl.rgb_to_xyz(test_rgb)
self.assertTuplesClose(test_xyz, colors['xyz'])
test_luv = husl.xyz_to_luv(test_xyz)
self.assertTuplesClose(test_luv, colors['luv'])
test_lch = husl.luv_to_lch(test_luv)
self.assertTuplesClose(test_lch, colors['lch'])
test_husl = husl.lch_to_husl(test_lch)
self.assertTuplesClose(test_husl, colors['husl'])
test_huslp = husl.lch_to_huslp(test_lch)
self.assertTuplesClose(test_huslp, colors['huslp'])
# Test backward functions
test_lch = husl.husl_to_lch(colors['husl'])
self.assertTuplesClose(test_lch, colors['lch'])
test_lch = husl.huslp_to_lch(colors['huslp'])
self.assertTuplesClose(test_lch, colors['lch'])
test_luv = husl.lch_to_luv(test_lch)
self.assertTuplesClose(test_luv, colors['luv'])
test_xyz = husl.luv_to_xyz(test_luv)
self.assertTuplesClose(test_xyz, colors['xyz'])
test_rgb = husl.xyz_to_rgb(test_xyz)
self.assertTuplesClose(test_rgb, colors['rgb'])
self.assertEqual(husl.rgb_to_hex(test_rgb), hex_color)
# Full test
self.assertEqual(husl.husl_to_hex(*colors['husl']), hex_color)
self.assertTuplesClose(husl.hex_to_husl(hex_color), colors['husl'])
self.assertEqual(husl.huslp_to_hex(*colors['huslp']), hex_color)
self.assertTuplesClose(husl.hex_to_huslp(hex_color),
colors['huslp'])
import matplotlib.pyplot as plt
import seaborn as sns
from palettable import cartocolors
from husl import hex_to_husl
colours = cartocolors.qualitative.Safe_10.hex_colors
cpalette = sns.color_palette(colours)
cpalette_light = sns.light_palette(hex_to_husl(colours[1]), input="husl")
cmap_qualitative = cartocolors.qualitative.Safe_10.mpl_colormap
cmap_light = sns.light_palette(hex_to_husl(colours[1]),
input="husl",
as_cmap=True)
cmap_diverging = sns.diverging_palette(h_neg=227,
h_pos=4,
s=73,
l=60,
n=12,
as_cmap=True)
cmap_diverging.set_bad('white')
pltmarkers = ['o', 'X', 'D', 'P', 'X']
def set_style():
sns.set_palette(cpalette)
sns.set_style(
'darkgrid', {
'axes.spines.right': False,
'axes.spines.top': False,
'axes.facecolor': '.93'
def plt_matches(data,
metric,
between='Generative Model',
by='Attack Classifier',
cols='Feature Set',
n=15):
between_groups = list(data.groupby(between).groups.keys())
combs = [c for c in combinations_with_replacement(between_groups, 2)]
if cols == 'Generative Model':
filtercmap = GMCMAP
elif cols == 'Attack Classifier':
filtercmap = ATTACKSCMAP
elif cols == 'Feature Set':
filtercmap = FEATURECMAP
else:
raise ValueError(f'Unknown groupby {between}')
plots = []
for fg, fgroup in data.groupby(cols):
groups = fgroup.groupby(by)
nrows = int(ceil(len(groups) / 2))
fig, ax = plt.subplots(nrows,
2,
figsize=(10, nrows * 4),
sharex=True,
sharey=True)
ax = ax.flatten()
for i, (g, group) in enumerate(groups):
subgroups = group.groupby(between)
pltdata = DataFrame(columns=between_groups, index=between_groups)
prey = {}
for sg, sgroup in subgroups:
prey[sg] = set(
sgroup.sort_values(metric)[['Target'
]].iloc[-n:].values.flatten())
for c in combs:
matches = len(prey[c[0]].intersection(prey[c[1]]))
pltdata.loc[c[0], c[1]] = matches
pltdata.loc[c[1], c[0]] = matches
try:
cmap = sns.light_palette(hex_to_husl(filtercmap[fg]),
input="husl",
as_cmap=True)
except:
cmap = cmap_light
sns.heatmap(pltdata,
ax=ax[i],
square=True,
annot=True,
vmin=0,
vmax=n,
cmap=cmap)
ax[i].set(title=g)
ax[i].set_xticklabels([x for x in ax[i].get_xticklabels()],
rotation=45,
ha='right')
if bool(len(groups) % 2):
ax[-1].axis('off')
fig.suptitle(
f'Matches between {n} most vulnerable targets by {by} for {cols} {fg}',
y=1.05)
plots.append(fig)
return plots
