def shCur(Me, Dev, ax=None, xs=[], labs=None, mkSiz=5, edWid=1, mks=None):
"""
Plot curve group.
Input
Me - mean matrix, k x n
k : #curves (algorithms)
n : #points (bins) per curve
Dev - standard variation matrix, k x n
ax - axis, {None}
xs - x position, {[]} | n x
mks - markers, {None} | {'o', 's'} | ...
parMk - marker parameter, {[]}
barWidG - width of group of histgram bar, {.8} <= 1
barWid - width of histgram bar in terms of 'barWidG / k', {.8} <= 1
devWid - width of deviation line in terms of the 'barWid', {1} <= 1
bdWid - width of boundary of bar, {1}
dev - dev flag, {'y'} | 'n'
yLim - threshold in y axis, {[]}
algs - algorithm names as legends, {[]} | 1 x nAlg (cell)
Output
ha - handle
"""
if ax is not None:
fig = plt.gcf()
fig.sca(ax)
# mk & cl
if mks is None:
mks, _ = genMkCl()
_, cls = genMkCl()
# devWid = ps(varargin, 'devWid', .1);
# bdWid = ps(varargin, 'bdWid', 1);
# parMk = ps(varargin, 'parMk', []);
# isDev = psY(varargin, 'dev', 'y');
# yLim = ps(varargin, 'yLim', []);
# algs = ps(varargin, 'algs', []);
# xs = ps(varargin, 'xs', []);
# dimension
k, n = Me.shape
# devWid = devWid * 1;
# x position
if len(xs) == 0:
xs = np.arange(1, n + 1)
# range
may = Me.max()
miy = Me.min()
# import pdb; pdb.set_trace()
# plot mean
haLns = []
for c in range(k):
cl = cls[c % len(cls)]
if labs is None:
ha, = plt.plot(xs, Me[c], '-' + mks[c], color=cl, markersize=mkSiz, markeredgewidth=edWid)
else:
ha, = plt.plot(xs, Me[c], '-' + mks[c], label=labs[c], color=cl, markersize=mkSiz, markeredgewidth=edWid)
haLns.append(ha)
return haLns
def shHst(mes, devs=[], ax=None, xs=[], barWid=0.8, devWid=0.8, bdWid=0, ori='ver', clG=None):
"""
Plot histogram.
Input
mes - histogram, 1 x n
devs - deviation, {[]} | 1 x n
ax - axis, {None}
xs - x position, {[]} | n x
barWid - width of histogram bar, {.8} <= 1
devWid - width of deviation line in terms of the 'barWid', {1} <= 1
bdWid - width of boundary of bar, {1}
ori - orientation, {'ver'} | 'hor'
clG - global color, {'None'} | 'red' | ...
lnWid - width of deviation line, {2}
dev - dev flag, {'y'} | 'n'
val - value flag, 'y' | {'n'}
form - value form, {'%d'}
yLim - threshold in y axis, {[]}
leg - legend, {[]}
Output
haBars
"""
if ax is not None:
fig = plt.gcf()
fig.sca(ax)
# mk & cl
mks, cls = genMkCl()
# dimension
n = len(mes)
# label position
if len(xs) == 0:
xs = np.arange(1, n + 1)
gap = xs[1] - xs[0] if n > 1 else .5
# width
barWid2 = 1.0 * barWid / 2 * gap
devWid2 = 1.0 * devWid / 2 * gap
# import pdb; pdb.set_trace()
# plot mean
haBars = cells(n)
for i in range(n):
x = xs[i]
y = mes[i]
xL = x - barWid2
xR = x + barWid2
if y == 0:
y = 1e-7
# continue
# color
if clG is None:
cl = cls[i % len(cls)]
else:
cl = clG
# draw
if bdWid == 0:
if ori == 'ver':
ha = plt.fill([xL, xR, xR, xL], [0, 0, y, y], color=cl, edgecolor=cl)
else:
ha = plt.fill([0, 0, y, y], [xL, xR, xR, xL], color=cl, edgecolor=cl)
else:
if ori == 'ver':
ha = plt.fill([xL, xR, xR, xL], [0, 0, y, y], color=cl, LineWidth=bdWid)
else:
ha = plt.fill([0, 0, y, y], [xL, xR, xR, xL], color=cl, LineWidth=bdWid)
haBars[i] = ha
# plot dev
if len(devs) > 0:
ys = mes + devs
for i in range(n):
x = xs(i)
y = mes(i)
dev = devs(i)
xL = x - devWid2
xR = x + devWid2
if y == 0:
continue
# color
cl = cls[i % len(cls)]
cl = 'k'
lnWid = 1
# vertical line
plt.plot([x, x], [-dev, dev] + y, 'Color', cl, 'LineWidth', lnWid)
# horizontal line
plt.plot([xL, xR], [dev, dev] + y, 'Color', cl, 'LineWidth', lnWid)
plt.plot([xL, xR], [-dev, -dev] + y, 'Color', cl, 'LineWidth', lnWid)
return haBars
def shHstG(Me, Dev=[], ax=None, xs=[], barWid=0.8, barWidG=.8, devWid=0.8, bdWid=0, ori='ver', clG=None):
"""
Plot histogram groups.
Input
Me - histogram, k x n
k: #bin in each group
n: #group
Dev - deviation, {[]} | 1 x n
ax - axis, {None}
xs - x position, {[]} | n x
barWid - width of histogram bar, {.8} <= 1
devWid - width of deviation line in terms of the 'barWid', {1} <= 1
bdWid - width of boundary of bar, {1}
ori - orientation, {'ver'} | 'hor'
clG - global color, {'None'} | 'red' | ...
lnWid - width of deviation line, {2}
dev - dev flag, {'y'} | 'n'
val - value flag, 'y' | {'n'}
form - value form, {'%d'}
yLim - threshold in y axis, {[]}
leg - legend, {[]}
Output
haBars
"""
# axis
if ax is not None:
fig = plt.gcf()
fig.sca(ax)
# mk & cl
mks, cls = genMkCl()
# dimension
k, n = Me.shape
# label position
if len(xs) == 0:
xs = np.arange(1, n + 1)
gap = xs[1] - xs[0] if n > 1 else .5
# width
barWidG2 = barWidG / 2
barWid = barWid * (barWidG / k)
gapWid = (barWidG - barWid * k) / (k - 1)
devWid2 = barWid * devWid / 2
# plot mean
HaBar = cells((k, n))
for i in range(n):
x = xs[i] - barWidG2
for c in range(k):
xL = x + (gapWid + barWid) * c
xR = xL + barWid
y = Me[c, i]
# color
if clG is None:
cl = cls[c % len(cls)]
else:
cl = clG
# draw
if bdWid == 0:
if ori == 'ver':
ha = plt.fill([xL, xR, xR, xL], [0, 0, y, y], color=cl, edgecolor=cl)
else:
ha = plt.fill([0, 0, y, y], [xL, xR, xR, xL], color=cl, edgecolor=cl)
else:
if ori == 'ver':
ha = plt.fill([xL, xR, xR, xL], [0, 0, y, y], color=cl, LineWidth=bdWid)
else:
ha = plt.fill([0, 0, y, y], [xL, xR, xR, xL], color=cl, LineWidth=bdWid)
HaBar[c, i] = ha
# plot dev
if Dev is not None:
ys = mes + devs
for i in range(n):
x = xs(i)
y = mes(i)
dev = devs(i)
xL = x - devWid2
xR = x + devWid2
if y == 0:
continue
# color
cl = cls[i % len(cls)]
cl = 'k'
lnWid = 1
# vertical line
plt.plot([x, x], [-dev, dev] + y, 'Color', cl, 'LineWidth', lnWid)
# horizontal line
plt.plot([xL, xR], [dev, dev] + y, 'Color', cl, 'LineWidth', lnWid)
plt.plot([xL, xR], [-dev, -dev] + y, 'Color', cl, 'LineWidth', lnWid)
return HaBar
def shCur(Me, Dev, ax=None, xs=[], labs=None, mkSiz=5, edWid=1, mks=None):
"""
Plot curve group.
Input
Me - mean matrix, k x n
k : #curves (algorithms)
n : #points (bins) per curve
Dev - standard variation matrix, k x n
ax - axis, {None}
xs - x position, {[]} | n x
mks - markers, {None} | {'o', 's'} | ...
parMk - marker parameter, {[]}
barWidG - width of group of histgram bar, {.8} <= 1
barWid - width of histgram bar in terms of 'barWidG / k', {.8} <= 1
devWid - width of deviation line in terms of the 'barWid', {1} <= 1
bdWid - width of boundary of bar, {1}
dev - dev flag, {'y'} | 'n'
yLim - threshold in y axis, {[]}
algs - algorithm names as legends, {[]} | 1 x nAlg (cell)
Output
ha - handle
"""
if ax is not None:
fig = plt.gcf()
fig.sca(ax)
# mk & cl
if mks is None:
mks, _ = genMkCl()
_, cls = genMkCl()
# devWid = ps(varargin, 'devWid', .1);
# bdWid = ps(varargin, 'bdWid', 1);
# parMk = ps(varargin, 'parMk', []);
# isDev = psY(varargin, 'dev', 'y');
# yLim = ps(varargin, 'yLim', []);
# algs = ps(varargin, 'algs', []);
# xs = ps(varargin, 'xs', []);
# dimension
k, n = Me.shape
# devWid = devWid * 1;
# x position
if len(xs) == 0:
xs = np.arange(1, n + 1)
# range
may = Me.max()
miy = Me.min()
# import pdb; pdb.set_trace()
# plot mean
haLns = []
for c in range(k):
cl = cls[c % len(cls)]
if labs is None:
ha, = plt.plot(xs,
Me[c],
'-' + mks[c],
color=cl,
markersize=mkSiz,
markeredgewidth=edWid)
else:
ha, = plt.plot(xs,
Me[c],
'-' + mks[c],
label=labs[c],
color=cl,
markersize=mkSiz,
markeredgewidth=edWid)
haLns.append(ha)
return haLns
def shHst(mes,
devs=[],
ax=None,
xs=[],
barWid=0.8,
devWid=0.8,
bdWid=0,
ori='ver',
clG=None):
"""
Plot histogram.
Input
mes - histogram, 1 x n
devs - deviation, {[]} | 1 x n
ax - axis, {None}
xs - x position, {[]} | n x
barWid - width of histogram bar, {.8} <= 1
devWid - width of deviation line in terms of the 'barWid', {1} <= 1
bdWid - width of boundary of bar, {1}
ori - orientation, {'ver'} | 'hor'
clG - global color, {'None'} | 'red' | ...
lnWid - width of deviation line, {2}
dev - dev flag, {'y'} | 'n'
val - value flag, 'y' | {'n'}
form - value form, {'%d'}
yLim - threshold in y axis, {[]}
leg - legend, {[]}
Output
haBars
"""
if ax is not None:
fig = plt.gcf()
fig.sca(ax)
# mk & cl
mks, cls = genMkCl()
# dimension
n = len(mes)
# label position
if len(xs) == 0:
xs = np.arange(1, n + 1)
gap = xs[1] - xs[0] if n > 1 else .5
# width
barWid2 = 1.0 * barWid / 2 * gap
devWid2 = 1.0 * devWid / 2 * gap
# import pdb; pdb.set_trace()
# plot mean
haBars = cells(n)
for i in range(n):
x = xs[i]
y = mes[i]
xL = x - barWid2
xR = x + barWid2
if y == 0:
y = 1e-7
# continue
# color
if clG is None:
cl = cls[i % len(cls)]
else:
cl = clG
# draw
if bdWid == 0:
if ori == 'ver':
ha = plt.fill([xL, xR, xR, xL], [0, 0, y, y],
color=cl,
edgecolor=cl)
else:
ha = plt.fill([0, 0, y, y], [xL, xR, xR, xL],
color=cl,
edgecolor=cl)
else:
if ori == 'ver':
ha = plt.fill([xL, xR, xR, xL], [0, 0, y, y],
color=cl,
LineWidth=bdWid)
else:
ha = plt.fill([0, 0, y, y], [xL, xR, xR, xL],
color=cl,
LineWidth=bdWid)
haBars[i] = ha
# plot dev
if len(devs) > 0:
ys = mes + devs
for i in range(n):
x = xs(i)
y = mes(i)
dev = devs(i)
xL = x - devWid2
xR = x + devWid2
if y == 0:
continue
# color
cl = cls[i % len(cls)]
cl = 'k'
lnWid = 1
# vertical line
plt.plot([x, x], [-dev, dev] + y, 'Color', cl, 'LineWidth', lnWid)
# horizontal line
plt.plot([xL, xR], [dev, dev] + y, 'Color', cl, 'LineWidth', lnWid)
plt.plot([xL, xR], [-dev, -dev] + y, 'Color', cl, 'LineWidth',
lnWid)
return haBars
def shHstG(Me,
Dev=[],
ax=None,
xs=[],
barWid=0.8,
barWidG=.8,
devWid=0.8,
bdWid=0,
ori='ver',
clG=None):
"""
Plot histogram groups.
Input
Me - histogram, k x n
k: #bin in each group
n: #group
Dev - deviation, {[]} | 1 x n
ax - axis, {None}
xs - x position, {[]} | n x
barWid - width of histogram bar, {.8} <= 1
devWid - width of deviation line in terms of the 'barWid', {1} <= 1
bdWid - width of boundary of bar, {1}
ori - orientation, {'ver'} | 'hor'
clG - global color, {'None'} | 'red' | ...
lnWid - width of deviation line, {2}
dev - dev flag, {'y'} | 'n'
val - value flag, 'y' | {'n'}
form - value form, {'%d'}
yLim - threshold in y axis, {[]}
leg - legend, {[]}
Output
haBars
"""
# axis
if ax is not None:
fig = plt.gcf()
fig.sca(ax)
# mk & cl
mks, cls = genMkCl()
# dimension
k, n = Me.shape
# label position
if len(xs) == 0:
xs = np.arange(1, n + 1)
gap = xs[1] - xs[0] if n > 1 else .5
# width
barWidG2 = barWidG / 2
barWid = barWid * (barWidG / k)
gapWid = (barWidG - barWid * k) / (k - 1)
devWid2 = barWid * devWid / 2
# plot mean
HaBar = cells((k, n))
for i in range(n):
x = xs[i] - barWidG2
for c in range(k):
xL = x + (gapWid + barWid) * c
xR = xL + barWid
y = Me[c, i]
# color
if clG is None:
cl = cls[c % len(cls)]
else:
cl = clG
# draw
if bdWid == 0:
if ori == 'ver':
ha = plt.fill([xL, xR, xR, xL], [0, 0, y, y],
color=cl,
edgecolor=cl)
else:
ha = plt.fill([0, 0, y, y], [xL, xR, xR, xL],
color=cl,
edgecolor=cl)
else:
if ori == 'ver':
ha = plt.fill([xL, xR, xR, xL], [0, 0, y, y],
color=cl,
LineWidth=bdWid)
else:
ha = plt.fill([0, 0, y, y], [xL, xR, xR, xL],
color=cl,
LineWidth=bdWid)
HaBar[c, i] = ha
# plot dev
if Dev is not None:
ys = mes + devs
for i in range(n):
x = xs(i)
y = mes(i)
dev = devs(i)
xL = x - devWid2
xR = x + devWid2
if y == 0:
continue
# color
cl = cls[i % len(cls)]
cl = 'k'
lnWid = 1
# vertical line
plt.plot([x, x], [-dev, dev] + y, 'Color', cl, 'LineWidth', lnWid)
# horizontal line
plt.plot([xL, xR], [dev, dev] + y, 'Color', cl, 'LineWidth', lnWid)
plt.plot([xL, xR], [-dev, -dev] + y, 'Color', cl, 'LineWidth',
lnWid)
return HaBar
