def anscombe():
"""
Creates 2x2 grid plot of the 4 anscombe datasets for illustration.
"""
_, ((axa, axb), (axc, axd)) = plt.subplots(2,
2,
sharex="col",
sharey="row")
colors = get_color_cycle()
for arr, ax, color in zip(ANSCOMBE, (axa, axb, axc, axd), colors):
x = arr[0]
y = arr[1]
# Set the X and Y limits
ax.set_xlim(0, 15)
ax.set_ylim(0, 15)
# Draw the points in the scatter plot
ax.scatter(x, y, c=color)
# Draw the linear best fit line on the plot
draw_best_fit(x, y, ax, c=color)
return (axa, axb, axc, axd)
def draw(self, y, y_pred):
"""
Parameters
----------
y : ndarray or Series of length n
An array or series of target or class values
y_pred : ndarray or Series of length n
An array or series of predicted target values
Returns
-------
ax : matplotlib Axes
The axis with the plotted figure
"""
# Some estimators particularly cross validation ones
# tend to provide choice to use different metrics for scoring,
# which we try to cater here
# If not available it falls back to the default score of R2.
try:
score_label = self.estimator.scoring
score_label = ' '.join(score_label.split('_')).capitalize()
except AttributeError:
score_label = "R2"
if score_label == "R2":
score_label = "$R^2$"
label = "{} $ = {:0.3f}$".format(score_label, self.score_)
self.ax.scatter(y,
y_pred,
c=self.colors["point"],
alpha=self.alpha,
label=label)
# TODO If score happens inside a loop, draw gets called multiple times.
# Ideally we'd want the best fit line to be drawn only once
if self.bestfit:
draw_best_fit(
y,
y_pred,
self.ax,
"linear",
ls="--",
lw=2,
c=self.colors["line"],
label="best fit",
)
# Set the axes limits based on the overall max/min values of
# concatenated X and Y data
# NOTE: shared_limits will be accounted for in finalize()
if self.shared_limits is True:
self.ax.set_xlim(min(min(y), min(y_pred)),
max(max(y), max(y_pred)))
self.ax.set_ylim(self.ax.get_xlim())
return self.ax
def datasaurus():
"""
Creates 2x2 grid plot of 4 from the Datasaurus Dozen datasets for illustration.
Citation:
Justin Matejka, George Fitzmaurice (2017)
Same Stats, Different Graphs: Generating Datasets with Varied Appearance and
Identical Statistics through Simulated Annealing
CHI 2017 Conference proceedings:
ACM SIGCHI Conference on Human Factors in Computing Systems
"""
_, ((axa, axb), (axc, axd)) = plt.subplots(2,
2,
sharex="col",
sharey="row")
colors = get_color_cycle()
for arr, ax, color in zip(DATASAURUS, (axa, axb, axc, axd), colors):
x = arr[0]
y = arr[1]
# Draw the points in the scatter plot
ax.scatter(x, y, c=color)
# Set the X and Y limits
ax.set_xlim(0, 100)
ax.set_ylim(0, 110)
# Draw the linear best fit line on the plot
draw_best_fit(x, y, ax, c=color)
return (axa, axb, axc, axd)
def datasaurus():
"""
Creates 2x2 grid plot of 4 from the Datasaurus Dozen datasets for illustration.
Citation:
Justin Matejka, George Fitzmaurice (2017)
Same Stats, Different Graphs: Generating Datasets with Varied Appearance and
Identical Statistics through Simulated Annealing
CHI 2017 Conference proceedings:
ACM SIGCHI Conference on Human Factors in Computing Systems
"""
_, ((axa, axb), (axc, axd)) = plt.subplots(2, 2, sharex='col', sharey='row')
colors = get_color_cycle()
for arr, ax, color in zip(DATASAURUS, (axa, axb, axc, axd), colors):
x = arr[0]
y = arr[1]
# Draw the points in the scatter plot
ax.scatter(x, y, c=color)
# Set the X and Y limits
ax.set_xlim(0, 100)
ax.set_ylim(0, 110)
# Draw the linear best fit line on the plot
draw_best_fit(x, y, ax, c=color)
return (axa, axb, axc, axd)
def draw(self, y, y_pred):
"""
Parameters
----------
y : ndarray or Series of length n
An array or series of target or class values
y_pred : ndarray or Series of length n
An array or series of predicted target values
Returns
-------
ax : matplotlib Axes
The axis with the plotted figure
"""
label = "$R^2 = {:0.3f}$".format(self.score_)
self.ax.scatter(y,
y_pred,
c=self.colors["point"],
alpha=self.alpha,
label=label)
# TODO If score happens inside a loop, draw gets called multiple times.
# Ideally we'd want the best fit line to be drawn only once
if self.bestfit:
draw_best_fit(
y,
y_pred,
self.ax,
"linear",
ls="--",
lw=2,
c=self.colors["line"],
label="best fit",
)
# Set the axes limits based on the range of X and Y data
# NOTE: shared_limits will be accounted for in finalize()
# TODO: do better than add one for really small residuals
self.ax.set_xlim(y.min() - 1, y.max() + 1)
self.ax.set_ylim(y_pred.min() - 1, y_pred.max() + 1)
return self.ax
def anscombe():
"""
Creates 2x2 grid plot of the 4 anscombe datasets for illustration.
"""
fig, ((axa, axb), (axc, axd)) = plt.subplots(2, 2, sharex='col', sharey='row')
for arr, ax in zip(ANSCOMBE, (axa, axb, axc, axd)):
x = arr[0]
y = arr[1]
# Set the X and Y limits
ax.set_xlim(0, 15)
ax.set_ylim(0, 15)
# Draw the points in the scatter plot
ax.scatter(x, y, c='g')
# Draw the linear best fit line on the plot
draw_best_fit(x, y, ax)
return (axa, axb, axc, axd)
def anscombe():
"""
Creates 2x2 grid plot of the 4 anscombe datasets for illustration.
"""
fig, ((axa, axb), (axc, axd)) = plt.subplots(2, 2, sharex='col', sharey='row')
colors = get_color_cycle()
for arr, ax, color in zip(ANSCOMBE, (axa, axb, axc, axd), colors):
x = arr[0]
y = arr[1]
# Set the X and Y limits
ax.set_xlim(0, 15)
ax.set_ylim(0, 15)
# Draw the points in the scatter plot
ax.scatter(x, y, c=color)
# Draw the linear best fit line on the plot
draw_best_fit(x, y, ax, c=color)
return (axa, axb, axc, axd)
def draw_joint(self, X, y, **kwargs):
"""
Draws the visualization for the joint axis.
"""
if self.joint_args is None:
self.joint_args = {}
self.joint_args.setdefault("alpha", 0.4)
facecolor = self.joint_args.pop("facecolor", "#dddddd")
self.joint_ax.set_facecolor(facecolor)
if self.joint_plot == "scatter":
aspect = self.joint_args.pop("aspect", "auto")
self.joint_ax.set_aspect(aspect)
self.joint_ax.scatter(X, y, **self.joint_args)
fit = self.joint_args.pop("fit", True)
if fit:
estimator = self.joint_args.pop("estimator", "linear")
draw_best_fit(X, y, self.joint_ax, estimator)
elif self.joint_plot == "hex":
x_bins = self.joint_args.pop("x_bins", 50)
y_bins = self.joint_args.pop("y_bins", 50)
colormap = self.joint_args.pop("cmap", 'Blues')
gridsize = int(np.mean([x_bins, y_bins]))
xmin = X.min()
xmax = X.max()
ymin = y.min()
ymax = y.max()
self.joint_ax.hexbin(X,
y,
gridsize=gridsize,
cmap=colormap,
mincnt=1,
**self.joint_args)
self.joint_ax.axis([xmin, xmax, ymin, ymax])
def draw_joint(self, X, y, **kwargs):
"""
Draws the visualization for the joint axis.
"""
if self.joint_args is None:
self.joint_args = {}
self.joint_args.setdefault("alpha", 0.4)
facecolor = self.joint_args.pop("facecolor", "#dddddd")
self.joint_ax.set_facecolor(facecolor)
if self.joint_plot == "scatter":
aspect = self.joint_args.pop("aspect", "auto")
self.joint_ax.set_aspect(aspect)
self.joint_ax.scatter(X, y, **self.joint_args)
fit = self.joint_args.pop("fit", True)
if fit:
estimator = self.joint_args.pop("estimator", "linear")
draw_best_fit(X, y, self.joint_ax, estimator)
elif self.joint_plot == "hex":
x_bins = self.joint_args.pop("x_bins", 50)
y_bins = self.joint_args.pop("y_bins", 50)
colormap = self.joint_args.pop("cmap", 'Blues')
gridsize = int(np.mean([x_bins, y_bins]))
xmin = X.min()
xmax = X.max()
ymin = y.min()
ymax = y.max()
self.joint_ax.hexbin(X, y,
gridsize=gridsize, cmap=colormap, mincnt=1, **self.joint_args
)
self.joint_ax.axis([xmin, xmax, ymin, ymax])
