Is there an easy way to avoid pyplot zooming far into noisy data?
Something like a lower boundary for the axis limits.
I am not trying to set a fix boundary to my axis, as this will fully disable automatic scaling.
Maybe a "minimum tick distance" would also work.
Right now I am using an additional 'invisible' plot in my graph that will define the maximum zoom.
Some example that illustrates what I want to achieve:
import numpy as np
import matplotlib.pyplot as plt
x = np.arange(0, 100, 1)
noise = np.random.randn(len(x))*0.1
y = 10+noise
y_dummy_low = [0]*len(x)
y_dummy_high = [20]*len(x)
plt.figure()
plt.plot(x, y) # noise data i actually want to plot
plt.plot(x, y_dummy_low, y_dummy_high, marker="None", linestyle="None") # this will avoid zooming too much
plt.show()
Zooming too far
Zooming OK
Related
I'm trying to zoom in a 3D plot. I'm using the ax.set_box_aspect() fonction. When doing so, the axis are zoomed in, they appear bigger, but the area where the data can be seen stay at the same size as before (the plot are not using the total available space).
The aim in the end is to have two axis, the first one 3d, the other one 2d. I would have wanted the first plot to take all the space available at the top half of the figure.
Here is the code before the Zoom
import numpy as np
import matplotlib.pyplot as plt
from mpl_toolkits.mplot3d import Axes3D
#---- generate data
nn = 100
X = np.random.randn(nn)*20 + 0
Y = np.random.randn(nn)*50 + 30
Z = np.random.randn(nn)*10 + -5
#---- check aspect ratio
asx, asy, asz = np.ptp(X), np.ptp(Y), np.ptp(Z)
fig = plt.figure(figsize=(15,15))
ax = fig.add_subplot(211, projection='3d')
#---- set box aspect ratio
ax.set_box_aspect((asx,asy,asz))
scat = ax.scatter(X, Y, Z, c=X+Y+Z, s=500, alpha=0.8)
ax.set_xlabel('X-axis'); ax.set_ylabel('Y-axis'); ax.set_zlabel('Z-axis')
ax = fig.add_subplot(212)
plt.show()
Before using the zoom
And now when I zoom in, the scatter is limitted in a square frame :
ax.set_box_aspect((asx,asy,asz), zoom = 2 )
After using the zoom
(The data used for the plot doesn't matter here, it is just to showcase my issue.)
I tried changing the axis limit with set_xlim3d or set_xlim, but in either case, the result is the same.
It seems like the showing area (I can't find the right word for it) stays a square no matter what.
I didn't find any usefull information on that matter online, (maybe from the lack of vocabulary to describe my problem).
I am trying to plot multiple rgb images with matplotlib
the code I am using is:
import numpy as np
import matplotlib.pyplot as plt
for i in range(0, images):
test = np.random.rand(1080, 720,3)
plt.subplot(images,2,i+1)
plt.imshow(test, interpolation='none')
the subplots appear tiny though as thumbnails
How can I make them bigger?
I have seen solutions using
fig, ax = plt.subplots()
syntax before but not with plt.subplot ?
plt.subplots initiates a subplot grid, while plt.subplot adds a subplot. So the difference is whether you want to initiate you plot right away or fill it over time. Since it seems, that you know how many images to plot beforehand, I would also recommend going with subplots.
Also notice, that the way you use plt.subplot you generate empy subplots in between the ones you are actually using, which is another reason they are so small.
import numpy as np
import matplotlib.pyplot as plt
images = 4
fig, axes = plt.subplots(images, 1, # Puts subplots in the axes variable
figsize=(4, 10), # Use figsize to set the size of the whole plot
dpi=200, # Further refine size with dpi setting
tight_layout=True) # Makes enough room between plots for labels
for i, ax in enumerate(axes):
y = np.random.randn(512, 512)
ax.imshow(y)
ax.set_title(str(i), fontweight='bold')
I have a vexing issue with a colorbar and even after vigorous research I cannot find the question even being asked. I have a plot where I overlay a contour and a pcolormesh and I would like a colorbar to indicate values. That works fine except for one thing:
The colorbar frame and color are offset
The colorbar frame and the actual bar are offset such that below you have a white bit in the frame and on top the color is poking out. While the frame is aligned with the axis as desired, the colorbar is offset.
Here is a working example that emulates the situation I was in, i.e. multiple plots with insets.
import matplotlib.gridspec as gridspec
import numpy as np
import matplotlib
import matplotlib.pyplot as plt
figheight = 4.2 - (2.1 - 49.519 / 25.4)
matplotlib.rcParams['figure.figsize'] = (5.25, figheight)
matplotlib.rcParams['axes.linewidth'] = 0.5
fig = plt.figure()
grid = gridspec.GridSpec(2, 1, height_ratios=[49.519 / 25.4 / figheight, 2.1 / figheight])
ax0 = plt.subplot(grid[0, 0])
ax1 = plt.subplot(grid[1, 0])
plt.tight_layout()
###############################################################################################
#
# Define position of inset
#
###############################################################################################
ax1.axis('off')
pos1 = ax1.get_position()
pos2 = matplotlib.transforms.Bbox([[pos1.x0, pos1.y0],
[.8*pos1.x1,
0.8*pos1.height + pos1.y0]])
left, bottom, width, height = [pos2.x0, pos2.y0, pos2.width, pos2.height]
ax2 = fig.add_axes([left, bottom, width, height])
###############################################################################################
#
# ax2 (inset) plot
#
###############################################################################################
pos2 = ax2.get_position()
ax2.axis('on')
x = np.linspace(0,5)
z = (np.outer(np.sin(x), np.cos(x))+1)*0.5
im = ax2.pcolormesh(z)
c = ax2.contour(z, linewidths=7)
ax2pos = ax2.get_position()
cbar_axis = fig.add_axes([ax2pos.x1+0.05,ax2pos.y0, .02, ax2pos.height])
colorbar = fig.colorbar(im, ax = ax2,
cax = cbar_axis, ticks = [0.1, .5, .9])
colorbar.outline.set_visible(True)
plot = 'Minimal.pdf'
fig.savefig(plot)
plt.close()
The problem persists in both the inline display and the saved .pdf if 'Inline' graphics backend is chosen. Using tight layout or not changes how badly the offset is depending on the size of the bar - same with using PyQT5 rather than inline graphics backend. I thought it was gone when I was changing between the various combinations, but I just realized it's still there.
I would appreciate any input.
As suggested by ImportanceOfBeingErnest I have tried using np.round on the figsize and that didn't change things. While you can fiddle around with sizes to make it look okay, it always stands over on one or the other side by some amount. When I change the graphics backend on Spyder 3 from 'Inline' to 'QT5' the problem becomes less severe with or without rounding. A summary of this is in this picture Colorbar overlap cases. Note that with not rounded and PyQT5 the problem still occurs, but is not as severe.
On inspection, it is clear that the colorbar is not only bleeding out over the top of its axes, but it's also positioned slightly to the left.
So, the problem here appears to be a conflict between the position of the colorbar axis and the colorbar itself when rasterization occurs. You can find more details on this issue in matplotlib's github repository, but I'll summarize what's going on here.
Colorbars are rasterized when the output is produced, so as to avoid artifacting issues during rendering. The position of the colorbar is snapped to the nearest integer pixels during the rasterization process, while the axis is kept where it is supposed to be. Then, when the output is produced, the colorbar falls within borders of fixed pixels of the image, despite the fact that the image is, itself, vectorized. Thus, there are two strategies that can be employed to avoid this mishap.
Use a finer DPI
The conversion from vectorized coordinates to rasterized coordinates takes place assuming a given DPI on the image. By default, this is set to be 72. However, by using more DPI, the overall shift induced by the rasterization process will be smaller, as the closest pixel the colorbar will snap to will be much nearer. Here, we change the output to have fig.savefig(plot,dpi=4000), and the problem goes away:
Note, however, that on my machine, the output size changed from 62 KB to 78 KB due to this change (although the DPI adjustment was also, admittedly, extreme). If you are worried about file sizes, you should pick a lower DPI that fixes the problem.
Use a different colormap
This rasterization happens when more than 50 colors are in the colorbar. Thus, we can do a quick test, setting our colormap to Pastel1 via
im = ax2.pcolormesh(z,cmap='Pastel1'). Here, the colorbar / axis mismatch is mitigated.
As a fallback, adopting a colorbar with fewer than 50 colors should mitigate this problem.
Rasterize the Axis
For completeness, there is also a third option. If you rasterize the colorbar axis, both the axis boundaries and the colormap will be rasterized, and you'll lose the offset. This will also rasterize your labels, and the axis will shift as one, breaking alignment with the nearby axis. For this, you just need to include cbar_axis.set_rasterized(True).
First, a way to overlay a contour and a pcolormesh and create a colorbar would be the following
import matplotlib.pyplot as plt
from mpl_toolkits.axes_grid1 import make_axes_locatable
import numpy as np
x = np.linspace(0,5)
z = (np.outer(np.sin(x), np.cos(x))+1)*0.5
fig = plt.figure(figsize=(4, 4))
ax = fig.add_subplot(111)
im = ax.pcolormesh(z)
c = ax.contour(z, linewidths=7)
divider = make_axes_locatable(ax)
cax = divider.append_axes("right", "5%", pad="3%")
colorbar = fig.colorbar(im, cax=cax, ticks = [0.1, .5, .9])
plt.show()
Now to the problem from the question. It is of course possible to create the axes to put the colorbar in manually. Replacing the colorbar creation with the code from the question still produces a nice image.
import matplotlib.pyplot as plt
import numpy as np
x = np.linspace(0,5)
z = (np.outer(np.sin(x), np.cos(x))+1)*0.5
fig = plt.figure(figsize=(4, 4))
ax = fig.add_subplot(111)
plt.subplots_adjust(right=0.8)
im = ax.pcolormesh(z)
c = ax.contour(z, linewidths=7)
ax2pos = ax.get_position()
cbar_axis = fig.add_axes([ax2pos.x1+0.05,ax2pos.y0, .05, ax2pos.height])
colorbar = fig.colorbar(im, ax = ax,
cax = cbar_axis, ticks = [0.1, .5, .9])
colorbar.outline.set_visible(True)
plt.show()
Conclusion so far: The issue is not reproducible, at least not without a Minimal, Complete, and Verifiable example.
I'm uncertain about the reasons for the behaviour in the example from the question. However, it seems that it can be overcome by rounding the figure size to 3 significant digits
matplotlib.rcParams['figure.figsize'] = (5.25, np.round(figheight,3))
I often need to indicate the distribution of some data in a concise plot, as in the below figure. I do this by plotting several fill_between areas, limited by the quantiles of the distribution.
ax.fill_between(x, quantile1, quantile2, alpha=0.2)
In a for loop, I make plots like this by calculating quantiles 1 and 2 (as indicated by the legend) as the 0% to 100% quantiles, then 10% to 90% and so on, each fill_between plotting on top of the previous "layer".
Here is the output with three layers of transparent colors along with the median line (0.5):
However, the legend colors are not what I would like them to be, since they (naturally) use the color of each individual layer, not taking into account the combined effect of several layers.
ax.legend([0.5]+[['0.0%', '100.0%'], ['10.0%', '90.0%'], ['30.0%', '70.0%']])
What is the best way to overwrite the face color value within the legend command?
I would like to avoid doing this by first plotting 0% to 10% with transparency "0.2", then 10% to 30% with transparency "0.4" and so on, as this will take twice the amount of time to compute and will make the code more complicated.
You can use proxy artists to place in the legend which have the exact same transparency as the resulting overlay from the plot.
As a proxy artist you can use a simple rectangle. The transparency however needs to be calculated as two objects with transparency 0.2 together will appear as a single object with transparency 0.36 (and not 0.4!).
import matplotlib.pyplot as plt
import numpy as np
import matplotlib.patches
a = np.sort(np.random.rand(6,18), axis=0)
x = np.arange(len(a[0]))
def alpha(i, base=0.2):
l = lambda x: x+base-x*base
ar = [l(0)]
for j in range(i):
ar.append(l(ar[-1]))
return ar[-1]
fig, ax = plt.subplots(figsize=(4,2))
handles = []
labels=[]
for i in range(len(a)/2):
ax.fill_between(x, a[i, :], a[len(a)-1-i, :], color="blue", alpha=0.2)
handle = matplotlib.patches.Rectangle((0,0),1,1,color="blue", alpha=alpha(i, base=0.2))
handles.append(handle)
label = "quant {:.1f} to {:.1f}".format(float(i)/len(a)*100, 100-float(i)/len(a)*100)
labels.append(label)
plt.legend(handles=handles, labels=labels, framealpha=1)
plt.show()
One has to decide if this is really worth the effort. A solution without transparency but with the very same result can be achieved much shorter:
import matplotlib.pyplot as plt
import numpy as np
a = np.sort(np.random.rand(6,18), axis=0)
x = np.arange(len(a[0]))
fig, ax = plt.subplots(figsize=(4,2))
for i in range(len(a)/2):
label = "quant {:.1f} to {:.1f}".format(float(i)/len(a)*100, 100-float(i)/len(a)*100)
c = plt.cm.Blues(0.2+.6*(float(i)/len(a)*2) )
ax.fill_between(x, a[i, :], a[len(a)-1-i, :], color=c, label=label)
plt.legend( framealpha=1)
plt.show()
Is there a way to automatically not display tick mark labels if they would protrude past the axis itself? For example, consider the following code
#!/usr/bin/python
import pylab as P, numpy as N, math as M
xvals=N.arange(-10,10,0.1)
yvals=[ M.sin(x) for x in xvals ]
P.plot( xvals, yvals )
P.show()
See how the -10 and 10 labels on the x-axis poke out to the left and right of the plot? And similar for the -1.0 and 1.0 labels on the y-axis. Can I automatically suppress plotting these but retain the ones that do not go outside the plot limits?
I think you could just format the axis ticks yourself and then prune the ones
that are hanging over. The recommended way to deal with setting up the axis is
to use the ticker API. So for example
from matplotlib.ticker import MaxNLocator
import matplotlib.pyplot as plt
import numpy as np
fig = plt.figure()
ax = fig.add_subplot(111)
xvals=np.arange(-10,10,0.1)
yvals=[ np.sin(x) for x in xvals ]
ax.plot( xvals, yvals )
ax.xaxis.set_major_locator(MaxNLocator(prune='both'))
plt.show()
Here we are creating a figure and axes, plotting the data, and then setting the xaxis
major ticks. The formatter MaxNLocator is given the
argument prune='both' which is described in the docs here.
This is not exactly what you were asking for, but maybe it will solve your problem.