plot changes not affecting all subplots [duplicate] - python

Im trying to plot a scatter matrix. I'm building on the example given in this thread Is there a function to make scatterplot matrices in matplotlib?. Here I have just modified the code slightly to make the axis visible for all the subplots. The modified code is given below
import itertools
import numpy as np
import matplotlib.pyplot as plt
def main():
np.random.seed(1977)
numvars, numdata = 4, 10
data = 10 * np.random.random((numvars, numdata))
fig = scatterplot_matrix(data, ['mpg', 'disp', 'drat', 'wt'],
linestyle='none', marker='o', color='black', mfc='none')
fig.suptitle('Simple Scatterplot Matrix')
plt.show()
def scatterplot_matrix(data, names, **kwargs):
"""Plots a scatterplot matrix of subplots. Each row of "data" is plotted
against other rows, resulting in a nrows by nrows grid of subplots with the
diagonal subplots labeled with "names". Additional keyword arguments are
passed on to matplotlib's "plot" command. Returns the matplotlib figure
object containg the subplot grid."""
numvars, numdata = data.shape
fig, axes = plt.subplots(nrows=numvars, ncols=numvars, figsize=(8,8))
fig.subplots_adjust(hspace=0.05, wspace=0.05)
for ax in axes.flat:
# Hide all ticks and labels
ax.xaxis.set_visible(True)
ax.yaxis.set_visible(True)
# # Set up ticks only on one side for the "edge" subplots...
# if ax.is_first_col():
# ax.yaxis.set_ticks_position('left')
# if ax.is_last_col():
# ax.yaxis.set_ticks_position('right')
# if ax.is_first_row():
# ax.xaxis.set_ticks_position('top')
# if ax.is_last_row():
# ax.xaxis.set_ticks_position('bottom')
# Plot the data.
for i, j in zip(*np.triu_indices_from(axes, k=1)):
for x, y in [(i,j), (j,i)]:
axes[x,y].plot(data[x], data[y], **kwargs)
# Label the diagonal subplots...
for i, label in enumerate(names):
axes[i,i].annotate(label, (0.5, 0.5), xycoords='axes fraction',
ha='center', va='center')
# Turn on the proper x or y axes ticks.
for i, j in zip(range(numvars), itertools.cycle((-1, 0))):
axes[j,i].xaxis.set_visible(True)
axes[i,j].yaxis.set_visible(True)
fig.tight_layout()
plt.xticks(rotation=45)
fig.show()
return fig
main()
I cant seem to be able to rotate the x-axis text of all the subplots. As it can be seen, i have tried the plt.xticks(rotation=45) trick. But this seems to perform the rotation for the last subplot alone.

Just iterate through the axes tied to the figure, set the active axes to the iterated object, and modify:
for ax in fig.axes:
matplotlib.pyplot.sca(ax)
plt.xticks(rotation=90)

plt only acts on the current active axes. You should bring it inside your last loop where you set some of the labels visibility to True:
# Turn on the proper x or y axes ticks.
for i, j in zip(range(numvars), itertools.cycle((-1, 0))):
axes[j,i].xaxis.set_visible(True)
axes[i,j].yaxis.set_visible(True)
for tick in axes[i,j].get_xticklabels():
tick.set_rotation(45)
for tick in axes[j,i].get_xticklabels():
tick.set_rotation(45)

for ax in fig.axes:
ax.tick_params(labelrotation=90)

Related

python code to place x and y label in the middle of tick positions in matplotlib.pyplot

I am trying to create a heatmap by putting gridlines to some particular positions which I have done. Suppose, I tried to make gridlines in positions 358 and 589 in a matrix of length 640,640. After that, I wanted to change the label from 358 to a defined value of 999 and 589 to a specified value of 1023. However, I cannot change the x and y labels in the center position of two gridlines. For example, I have tried the following:
data = np.random.rand(640, 640)
fig, ax = plt.subplots()
im = ax.imshow(data,cmap='coolwarm')
ax.set_xticks([358,589])
ax.set_yticks([358,589])
ax.set_xticklabels([999,1023])
ax.set_yticklabels([999,1023])
ax.grid(which='major',color='black',linestyle='--',linewidth=1,alpha=0.5)
plt.show()
That create a image as follows:
Heatmap with customized labelling
But I want the labeling in the middle of two gridlines instead of the gridline positions. How can that be done?
By default, both the tick labels and the grid lines are decided via the major ticks. To change this, you could use the minor ticks to position the grid lines and the major ticks for the tick labels:
from matplotlib import pyplot as plt
import numpy as np
data = np.random.randn(640, 640).cumsum(axis=0).cumsum(axis=1)
fig, ax = plt.subplots()
im = ax.imshow(data, cmap='coolwarm')
positions = np.array([358, 589])
ax.set_xticks(positions, minor=True)
ax.set_yticks(positions, minor=True)
borders = np.append(0, positions)
mids = (borders[:-1] + borders[1:]) / 2
ax.set_xticks(mids, [999, 1023], minor=False)
ax.set_yticks(mids, [999, 1023], minor=False)
ax.grid(which='minor', color='black', linestyle='--', linewidth=1, alpha=0.9)
plt.show()

Resize axes of top and right joint marginal plots to match central plot with matplotlib

How do you size the axes of a marginal plot to match the size of a non-square central plot using matplotlib?
In the image, you'll see that the top marginal plot is too wide, even though it shares the x-axis labels.
Context: I'm trying to create a joint plot like in Seaborn, but with a non-square heatmap at center and bar graphs as the marginal plots. JointGrids isn't designed to work with heatmaps (which is okay, on to matplotlib!). Merging a matplotlib heatmap with subplot barplots gets me close, but I find one bargraph's axis is larger than the central heatmap even when I share axes.
Minimum working example:
import numpy as np
import matplotlib.pyplot as plt
import pandas as pd
### Function from matplotlib ################################################
def heatmap(data, row_labels, col_labels, ax=None,
cbar_kw={}, cbarlabel="", **kwargs):
"""
Create a heatmap from a numpy array and two lists of labels.
Parameters
----------
data
A 2D numpy array of shape (M, N).
row_labels
A list or array of length M with the labels for the rows.
col_labels
A list or array of length N with the labels for the columns.
ax
A `matplotlib.axes.Axes` instance to which the heatmap is plotted. If
not provided, use current axes or create a new one. Optional.
cbar_kw
A dictionary with arguments to `matplotlib.Figure.colorbar`. Optional.
cbarlabel
The label for the colorbar. Optional.
**kwargs
All other arguments are forwarded to `imshow`.
"""
if not ax:
ax = plt.gca()
# Plot the heatmap
im = ax.imshow(data, **kwargs)
# Create colorbar
cbar = ax.figure.colorbar(im, ax=ax, **cbar_kw)
cbar.ax.set_ylabel(cbarlabel, rotation=-90, va="bottom")
# Show all ticks and label them with the respective list entries.
ax.set_xticks(np.arange(data.shape[1]), labels=col_labels)
ax.set_yticks(np.arange(data.shape[0]), labels=row_labels)
# Rotate the tick labels and set their alignment.
plt.setp(ax.get_xticklabels(), rotation=-30, ha="right",
rotation_mode="anchor")
# Turn spines off and create white grid.
ax.spines[:].set_visible(False)
ax.set_xticks(np.arange(data.shape[1]+1)-.5, minor=True)
ax.set_yticks(np.arange(data.shape[0]+1)-.5, minor=True)
ax.grid(which="minor", color="w", linestyle='-', linewidth=3)
ax.tick_params(which="minor", bottom=False, left=False)
return im, cbar
### Now this specific case ################################################
# Make dataframe
dd = {f'Col_{col}': [col*x**2 for x in range(25)] for col in range(16)}
index = [f'Row_{x}' for x in range(25)]
df = pd.DataFrame(dd, index=index)
# Make means by axis
ax0_means = df.mean(axis=0)
ax1_means = df.mean(axis=1)
# Build figure and axes
fig, axs = plt.subplots(2, 2, sharex="col", sharey="row", figsize=(16,16),
gridspec_kw=dict(height_ratios=[1, 3],width_ratios=[3, 1]))
axs[0, 1].set_visible(False)
axs[0, 0].set_box_aspect(1/3)
axs[1, 1].set_box_aspect(3/1)
# Plot data
im, cbar = heatmap(df, df.index, df.columns, ax=axs[1,0])
plt.setp(axs[1,0].get_xticklabels(), rotation=45, ha="right",
rotation_mode="anchor")
# Rotate the tick labels and set their alignment.
axs[1, 1].barh(y=ax1_means.index, width=ax1_means.values)
axs[0, 0].bar(x=ax0_means.index, height=ax0_means.values)
plt.show()
As the heatmap gets a default "equal" aspect ratio, and gets shrunk due to the colorbar, an idea is to manually resize the histograms once everything is created.
from matplotlib.transforms import Bbox
# code added at the end, just before plt.show()
(x0m, y0m), (x1m, y1m) = axs[1, 0].get_position().get_points() # main heatmap
(x0h, y0h), (x1h, y1h) = axs[0, 0].get_position().get_points() # horizontal histogram
axs[0, 0].set_position(Bbox([[x0m, y0h], [x1m, y1h]]))
(x0v, y0v), (x1v, y1v) = axs[1, 1].get_position().get_points() # vertical histogram
axs[1, 1].set_position(Bbox([[x0v, y0m], [x1v, y1m]]))
plt.show()
(The following example uses hspace=0.01, wspace=0.02 in the gridspec_kw=)

Is there a way to replace a matplotlib subplot with a legend (rather than have the legend outside the subplots)?

I have a figure with 11 scatter plots as subplots. I would like the legend (same across all 11 subplots) to replace the 12th subplot. Is there a way to put the legend there and have it be the same size as the subplots?
Matplotlib scatter plot of 11 subplots
Sort of a manual approach, but here it is:
You can "remove" an axis using ax.clear() and ax.set_axis_off(). Then you can create patches with specific colors and labels, and create a legend in the desired ax based on them.
Try this:
import matplotlib.pyplot as plt
import matplotlib.patches as mpatches
import numpy as np
# Create figure with subplots
fig, axes = plt.subplots(figsize=(16, 16), ncols=4, nrows=3, sharex=True, sharey=True)
# Plot some random data
for row in axes:
for ax in row:
ax.scatter(np.random.random(5), np.random.random(5), color='green')
ax.scatter(np.random.random(2), np.random.random(2), color='red')
ax.scatter(np.random.random(3), np.random.random(3), color='orange')
ax.set_title('some title')
# Clear bottom-right ax
bottom_right_ax = axes[-1][-1]
bottom_right_ax.clear() # clears the random data I plotted previously
bottom_right_ax.set_axis_off() # removes the XY axes
# Manually create legend handles (patches)
red_patch = mpatches.Patch(color='red', label='Red data')
green_patch = mpatches.Patch(color='green', label='Green data')
orange_patch = mpatches.Patch(color='orange', label='Orange data')
# Add legend to bottom-right ax
bottom_right_ax.legend(handles=[red_patch, green_patch, orange_patch], loc='center')
# Show figure
plt.show()
Output:

How to align logarithmic scale ticks across subplots?

I want to fix the position of the ticks on the logarithmic scale, such that they are the same in each subplot (see red annotation in image).
My code looks like this:
ax = fig.add_subplot(2,2, axis)
ax2 = ax.twinx()
ax2.set_yscale('log')
ax2.set_ylim(0,100)
Right now, set_yscale=('log') optimizes the tick spacing for each subplot. I prefer to adopt the tick spacing of the upper right subplot.
You can achieve this by getting the limits of the left twin axis and setting it as the limits of the right twin axis.
Consider the following working example. Follow this procedure for the subplots you want to align the axes of.
import numpy as np
import matplotlib.pyplot as plt
fig = plt.figure(figsize=(8, 3))
axl = fig.add_subplot(121)
axr = fig.add_subplot(122)
ax1 = axl.twinx()
ax1.plot(np.logspace(-2, 3, 5))
ax1.set_yscale('log')
ax2 = axr.twinx()
ax2.plot(np.logspace(0, 3, 5))
ax2.set_yscale('log')
ax2.set_ylim(ax1.get_ylim()) # <-- This is the key line
plt.tight_layout()
plt.show()
OP's solution:
Plot a dummy curve and set alpha=0. Make sure the curve spans y_min and y_max.
fig = plt.figure()
axes = [1,2,3,4]
for axis in axes:
ax = fig.add_subplot(2,2, axis)
ax2 = ax.twinx()
ax2.set_yscale('log')
ax2.plot(x_dummy, y_dummy, alpha=0) # <-- dummy plot
x_real, y_real = func_that_loads_data() # <-- your interesting plot
curve1 = ax2.plot(x_real, y_real)
plt.show()
The solution provided by Sheldore was impractical to implement because I plot my data using a for-loop (unavoidable unless I escalate the number of variables).
Since I overwrite the ax variable on every iteration, I would have to save the y-limit as a global variable. Read here why global variables should be avoided.
ax = fig.add_subplot(2,2, axis)
ax2 = ax.twinx()
ax2.set_yscale('log')
if axis == 1:
global yscale
yscale = ax2.get_ylim() # <-- where the magic happens
elif axis > 1:
ax2.set_ylim(yscale)

Tick properties for scatterplot matrices with Matplotlib

I am trying to plot a scatterplot matrix based on the code written by Joe Kington: Is there a function to make scatterplot matrices in matplotlib?
Some people already helped me: Thank you again (especially J.K.).
I am having a last problem: I cannot rotate the ticks of some axis for which numbers overlap (bottom left):
I would like to try to have them vertical but I cannot do it.... Here is my code:
import itertools
import numpy as np
import pylab as plot
import scipy
import matplotlib
import matplotlib.pyplot as plt
from matplotlib import axis
import math
from matplotlib import rc
import os
import platform
def main():
FigSize=8.89
FontSize=8
np.random.seed(1977)
numvars, numdata = 4, 10
data = 10 * np.random.random((numvars, numdata))
fig = scatterplot_matrix(data, ['mpg', 'disp', 'drat', 'wt'], FigSize, FontSize,
linestyle='none', marker='o', color='black', mfc='none', markersize=3,)
fig.suptitle('Simple Scatterplot Matrix')
plt.savefig('Plots/ScatterplotMatrix/ScatterplotMatrix2.pdf',format='pdf', dpi=1000, transparent=True, bbox_inches='tight')
plt.show()
def scatterplot_matrix(data, names, FigSize, FontSize, **kwargs):
"""Plots a scatterplot matrix of subplots. Each row of "data" is plotted
against other rows, resulting in a nrows by nrows grid of subplots with the
diagonal subplots labeled with "names". Additional keyword arguments are
passed on to matplotlib's "plot" command. Returns the matplotlib figure
object containg the subplot grid."""
legend=['(kPa)','\%','\%','\%']
numvars, numdata = data.shape
fig, axes = plt.subplots(nrows=numvars, ncols=numvars, figsize=(FigSize/2.54,FigSize/2.54))
fig.subplots_adjust(hspace=0.05, wspace=0.05)
sub_labelx_top=[2,4]
sub_labelx_bottom=[13,15]
sub_labely_left=[5,13]
sub_labely_right=[4,12]
for i, ax in enumerate(axes.flat, start=1):
# Hide all ticks and labels
ax.xaxis.set_visible(False)
ax.yaxis.set_visible(False)
ax.xaxis.set_major_locator(MaxNLocator(prune='both',nbins=4))
ax.yaxis.set_major_locator(MaxNLocator(prune='both',nbins=4)) #http://matplotlib.org/api/ticker_api.html#matplotlib.ticker.MaxNLocator
# Set up ticks only on one side for the "edge" subplots...
if ax.is_first_col():
ax.yaxis.set_ticks_position('left')
ax.tick_params(direction='out')
ax.yaxis.set_tick_params(labelsize=0.75*FontSize)
if i in sub_labely_left:
ax.yaxis.set_label_position('left')
ax.set_ylabel('(\%)',fontsize=0.75*FontSize)
if ax.is_last_col():
ax.yaxis.set_ticks_position('right')
ax.tick_params(direction='out')
ax.yaxis.set_tick_params(labelsize=0.75*FontSize)
if i in sub_labely_right:
ax.yaxis.set_label_position('right')
if i==4:
ax.set_ylabel('(kPa)',fontsize=0.75*FontSize)
else:
ax.set_ylabel('(\%)',fontsize=0.75*FontSize)
if ax.is_first_row():
ax.xaxis.set_ticks_position('top')
ax.tick_params(direction='out')
ax.xaxis.set_tick_params(labelsize=0.75*FontSize)
if i in sub_labelx_top:
ax.xaxis.set_label_position('top')
ax.set_xlabel('(\%)',fontsize=0.75*FontSize)
if ax.is_last_row():
ax.xaxis.set_ticks_position('bottom')
ax.tick_params(direction='out')
ax.xaxis.set_tick_params(labelsize=0.75*FontSize)
if i in sub_labelx_bottom:
ax.xaxis.set_label_position('bottom')
if i==13:
ax.set_xlabel('(kPa)',fontsize=0.75*FontSize)
else:
ax.set_xlabel('(\%)',fontsize=0.75*FontSize)
# Plot the data.
for i, j in zip(*np.triu_indices_from(axes, k=1)):
for x, y in [(i,j), (j,i)]:
axes[x,y].plot(data[y], data[x], **kwargs)
# Label the diagonal subplots...
for i, label in enumerate(names):
axes[i,i].annotate(label, (0.5, 0.5), xycoords='axes fraction',
ha='center', va='center',fontsize=FontSize)
# Turn on the proper x or y axes ticks.
for i, j in zip(range(numvars), itertools.cycle((-1, 0))):
axes[j,i].xaxis.set_visible(True)
axes[i,j].yaxis.set_visible(True)
return fig
main()
My second question is more for the 'fun': how can I make the subplots perfectly squares?
I apologize to Joe Kington; I know my code is way less elegant than his... I just started few weeks ago. If you have any suggestions to improve mine, for example to make it more dynamic, I am very interesting.
You can rotate the xtick labels using setp.
from matplotlib.artist import setp
Then after you set the x tick positions for the top row and left column of subplot call:
setp(ax.get_xticklabels(), rotation=90)
To make the size of the subplots equal, you can fig.subplots_adjust to set the area of all the subplots to a square. Something like this:
gridSize = 0.6
leftBound = 0.5 - gridSize/2
bottomBound = 0.1
rightBound = leftBound + gridSize
topBound = bottomBound + gridSize
fig.subplots_adjust(hspace=0.05, wspace=0.05, left=leftBound,
bottom=bottomBound, right=rightBound, top=topBound)
If the figure size isn't square, you'll need to change the shape of the grid accordingly. Alternately, you could add each subplot axes individually with fig.add_axes. That will allow you to set the size directly but you'll also have to set the location.
Don't use bbox_inches='tight' to save the figure or you'll lose the title with these setting. You can save like this:
plt.savefig('ScatterplotMatrix.pdf',format='pdf', dpi=1000, transparent=True)
The resulting graph looks like this:

Categories

Resources