def distplot(a,
bins=None,
hist=True,
kde=True,
rug=False,
fit=None,
hist_kws=None,
kde_kws=None,
rug_kws=None,
fit_kws=None,
color=None,
vertical=False,
xlabel=None,
ax=None):
"""Flexibly plot a distribution of observations.
Parameter
a : (squeezable to) 1d array
Observed data.
bins : argument for matplotlib hist(), or None
Specification of hist bins, or None to use Freedman-Diaconis rule.
hist : bool, default True
Whether to plot a (normed) histogram.
kde : bool, default True
Whether to plot a gaussian kernel density estimate.
rug : bool, default False
Whether to draw a rugplot on the support axis.
fit : random variable object
An object with `fit` method, returning a tuple that can be passed to a
`pdf` method a positional arguments following an grid of values to
evaluate the pdf on.
{hist, kde, rug, fit}_kws : dictionaries
Keyword arguments for underlying plotting functions.
color : matplotlib color, optional
Color to plot everything but the fitted curve in.
vertical : bool, default False
If True, oberved values are on y-axis.
xlabel : string, False, or None
Name for the x axis label. if None, will try to get it from a.name
if False, do not set the x label.
ax : matplotlib axis, optional
if provided, plot on this axis
Returns
-------
ax : matplotlib axis
"""
if ax is None:
ax = plt.gca()
# Intelligently label the axis
label_x = bool(xlabel)
if xlabel is None and hasattr(a, "name"):
xlabel = a.name
if xlabel is not None:
label_x = True
# Make a a 1-d array
a = np.asarray(a).squeeze()
# Handle dictionary defaults
if hist_kws is None:
hist_kws = dict()
if kde_kws is None:
kde_kws = dict()
if rug_kws is None:
rug_kws = dict()
if fit_kws is None:
fit_kws = dict()
# Get the color from the current color cycle
if color is None:
if vertical:
line, = ax.plot(0, a.mean())
else:
line, = ax.plot(a.mean(), 0)
color = line.get_color()
line.remove()
if hist:
if bins is None:
# From http://stats.stackexchange.com/questions/798/
h = 2 * moss.iqr(a) * len(a)**-(1 / 3)
bins = (a.max() - a.min()) / h
hist_alpha = hist_kws.pop("alpha", 0.4)
orientation = "horizontal" if vertical else "vertical"
hist_color = hist_kws.pop("color", color)
ax.hist(a,
bins,
normed=True,
color=hist_color,
alpha=hist_alpha,
orientation=orientation,
**hist_kws)
if kde:
kde_color = kde_kws.pop("color", color)
kdeplot(a, vertical=vertical, color=kde_color, ax=ax, **kde_kws)
if rug:
rug_color = rug_kws.pop("color", color)
axis = "y" if vertical else "x"
rugplot(a, axis=axis, color=rug_color, ax=ax, **rug_kws)
if fit is not None:
fit_color = fit_kws.pop("color", "#282828")
npts = fit_kws.pop("npts", 1000)
support_thresh = fit_kws.pop("support_thresh", 1e-4)
params = fit.fit(a)
pdf = lambda x: fit.pdf(x, *params)
x = _kde_support(a, pdf, npts, support_thresh)
y = pdf(x)
if vertical:
x, y = y, x
ax.plot(x, y, color=fit_color, **fit_kws)
if label_x:
ax.set_xlabel(xlabel)
return ax
def _freedman_diaconis_bins(a):
"""Calculate number of hist bins using Freedman-Diaconis rule."""
# From http://stats.stackexchange.com/questions/798/
a = np.asarray(a)
h = 2 * moss.iqr(a) / (len(a) ** (1 / 3))
return np.ceil((a.max() - a.min()) / h)
def distplot(a, bins=None, hist=True, kde=True, rug=False, fit=None,
hist_kws=None, kde_kws=None, rug_kws=None, fit_kws=None,
color=None, vertical=False, legend=False, xlabel=None, ax=None):
"""Flexibly plot a distribution of observations.
Parameters
----------
a : (squeezable to) 1d array
observed data
bins : argument for matplotlib hist(), or None
specification of bins or None to use Freedman-Diaconis rule
hist : bool, default True
whether to plot a (normed) histogram
kde : bool, defualt True
whether to plot a gaussian kernel density estimate
rug : bool, default False
whether to draw a rugplot on the support axis
fit : random variable object
object with `fit` method returning a tuple that can be
passed to a `pdf` method a positional arguments following
an array of values to evaluate the pdf at
{hist, kde, rug, fit}_kws : dictionaries
keyword arguments for underlying plotting functions
color : matplotlib color, optional
color to plot everything but the fitted curve in
vertical : bool, default False
if True, oberved values are on y-axis
legend : bool, default True
if True, add a legend to the plot with what the plotted lines are
xlabel : string, False, or None
name for the x axis label. if None, will try to get it from a.name
if False, do not set the x label
ax : matplotlib axis, optional
if provided, plot on this axis
Returns
-------
ax : matplotlib axis
"""
if ax is None:
ax = plt.gca()
# Intelligently label the axis
label_x = bool(xlabel)
if xlabel is None and hasattr(a, "name"):
xlabel = a.name
if xlabel is not None:
label_x = True
# Make a a 1-d array
a = np.asarray(a).squeeze()
# Handle dictionary defaults
if hist_kws is None:
hist_kws = dict()
if kde_kws is None:
kde_kws = dict()
if rug_kws is None:
rug_kws = dict()
if fit_kws is None:
fit_kws = dict()
# Get the color from the current color cycle
if color is None:
if vertical:
line, = ax.plot(0, a.mean())
else:
line, = ax.plot(a.mean(), 0)
color = line.get_color()
line.remove()
if hist:
if bins is None:
# From http://stats.stackexchange.com/questions/798/
h = 2 * moss.iqr(a) * len(a) ** -(1 / 3)
bins = (a.max() - a.min()) / h
hist_alpha = hist_kws.pop("alpha", 0.4)
orientation = "horizontal" if vertical else "vertical"
hist_color = hist_kws.pop("color", color)
ax.hist(a, bins, normed=True, color=hist_color, alpha=hist_alpha,
orientation=orientation, **hist_kws)
if kde:
kde_color = kde_kws.pop("color", color)
kde_kws["label"] = "kde"
kdeplot(a, vertical=vertical, color=kde_color, ax=ax, **kde_kws)
if rug:
rug_color = rug_kws.pop("color", color)
axis = "y" if vertical else "x"
rugplot(a, axis=axis, color=rug_color, ax=ax, **rug_kws)
if fit is not None:
fit_color = fit_kws.pop("color", "#282828")
npts = fit_kws.pop("npts", 1000)
support_thresh = fit_kws.pop("support_thresh", 1e-4)
params = fit.fit(a)
pdf = lambda x: fit.pdf(x, *params)
x = _kde_support(a, pdf, npts, support_thresh)
y = pdf(x)
if vertical:
x, y = y, x
fit_kws["label"] = fit.name
ax.plot(x, y, color=fit_color, **fit_kws)
if legend:
ax.legend(loc="best")
if label_x:
ax.set_xlabel(xlabel)
return ax
def distplot(a, bins=None, hist=True, kde=True, rug=False, fit=None,
hist_kws=None, kde_kws=None, rug_kws=None, fit_kws=None,
color=None, vertical=False, axlabel=None, ax=None):
"""Flexibly plot a distribution of observations.
Parameter
a : (squeezable to) 1d array
Observed data.
bins : argument for matplotlib hist(), or None
Specification of hist bins, or None to use Freedman-Diaconis rule.
hist : bool, default True
Whether to plot a (normed) histogram.
kde : bool, default True
Whether to plot a gaussian kernel density estimate.
rug : bool, default False
Whether to draw a rugplot on the support axis.
fit : random variable object
An object with `fit` method, returning a tuple that can be passed to a
`pdf` method a positional arguments following an grid of values to
evaluate the pdf on.
{hist, kde, rug, fit}_kws : dictionaries
Keyword arguments for underlying plotting functions.
color : matplotlib color, optional
Color to plot everything but the fitted curve in.
vertical : bool, default False
If True, oberved values are on y-axis.
axlabel : string, False, or None
Name for the support axis label. If None, will try to get it
from a.namel if False, do not set a label.
ax : matplotlib axis, optional
if provided, plot on this axis
Returns
-------
ax : matplotlib axis
"""
if ax is None:
ax = plt.gca()
# Intelligently label the support axis
label_ax = bool(axlabel)
if axlabel is None and hasattr(a, "name"):
axlabel = a.name
if axlabel is not None:
label_ax = True
# Make a a 1-d array
a = np.asarray(a).squeeze()
# Handle dictionary defaults
if hist_kws is None:
hist_kws = dict()
if kde_kws is None:
kde_kws = dict()
if rug_kws is None:
rug_kws = dict()
if fit_kws is None:
fit_kws = dict()
# Get the color from the current color cycle
if color is None:
if vertical:
line, = ax.plot(0, a.mean())
else:
line, = ax.plot(a.mean(), 0)
color = line.get_color()
line.remove()
if hist:
if bins is None:
# From http://stats.stackexchange.com/questions/798/
h = 2 * moss.iqr(a) * len(a) ** -(1 / 3)
bins = (a.max() - a.min()) / h
hist_alpha = hist_kws.pop("alpha", 0.4)
orientation = "horizontal" if vertical else "vertical"
hist_color = hist_kws.pop("color", color)
ax.hist(a, bins, normed=True, color=hist_color, alpha=hist_alpha,
orientation=orientation, **hist_kws)
if kde:
kde_color = kde_kws.pop("color", color)
kdeplot(a, vertical=vertical, color=kde_color, ax=ax, **kde_kws)
if rug:
rug_color = rug_kws.pop("color", color)
axis = "y" if vertical else "x"
rugplot(a, axis=axis, color=rug_color, ax=ax, **rug_kws)
if fit is not None:
fit_color = fit_kws.pop("color", "#282828")
gridsize = fit_kws.pop("gridsize", 500)
cut = fit_kws.pop("cut", 3)
clip = fit_kws.pop("clip", (-np.inf, np.inf))
bw = sm.nonparametric.bandwidths.bw_scott(a)
x = _kde_support(a, bw, gridsize, cut, clip)
params = fit.fit(a)
pdf = lambda x: fit.pdf(x, *params)
y = pdf(x)
if vertical:
x, y = y, x
ax.plot(x, y, color=fit_color, **fit_kws)
if label_ax:
if vertical:
ax.set_ylabel(axlabel)
else:
ax.set_xlabel(axlabel)
return ax
def _freedman_diaconis_bins(a):
"""Calculate number of hist bins using Freedman-Diaconis rule."""
# From http://stats.stackexchange.com/questions/798/
a = np.asarray(a)
h = 2 * moss.iqr(a) / (len(a)**(1 / 3))
return np.ceil((a.max() - a.min()) / h)