I have noticed a frustrating problem with Pythons Matplotlib where matrix plotting produces
an uneven grid. This issue is persistent with and without high DPI, as well as in EPS files.
The following code is used for the image generation:
import matplotlib.pyplot as plt
import numpy as np
arr = np.zeros((200,200))
# Set the diagonal to 1
arr[np.arange(200), np.arange(200)] = 1
plt.matshow(arr)
plt.savefig('matshow_test.png', dpi=1000)
DPI=1000:
Which has the sizes 65x65, 90x90, 95x95, 90x90, 95x95 and so on.
DPI=default
Which varies between 1x1 and 2x2 for each cell.
EPS rendered in latex:
Which is clearly distorted.
My questions are:
Why is this the default behaviour of Matplotlib?
How can I fix this?
Using Python 3.9.10 with Matplotlib 3.5.1
The matplotlib function matshow uses an antialiasing filter on the images. Unfortunately it is enabled even for vector graphic backends such as (e)ps, pdf or svg. That means, the image is rasterized, antialiased to a specific size and than inlined in the vector graphic.
Antialiasing takes into account a specific display resolution (dpi) and image size. If you change those parameters when viewing an image (for example when zooming in) the image can get heavily distorted, as you have experienced.
There is a discussion about the default antialiasing for matplotlib imshow (and also matshow which uses the same mechanism) here.
You should be able to fix your issue (and get true vector graphics) by disabling the antialiasing with the
matshow(..., interpolation='none')
option.
Related
I am currently trying to visualize the phase of an electromagnetic field which is 2pi-periodic. To visualize that e.g. 1.9 pi is almost the same as 0, I am using a cyclic colormap (twilight). However, when I plot my images, there are always lines at the sections where the phase jumps from (almost) 2pi to 0. When you zoom in on these lines, these artefacts vanish.
Here is a simple script and example images that demonstrate this issue.
import numpy as np
import matplotlib.pyplot as plt
x = np.linspace(-3,3,501)
x,y = np.meshgrid(x,x)
data = x**2+y**2
data = np.mod(data, 2)
plt.set_cmap('twilight')
plt.imshow(data)
plt.show()
I tested it with "twilight_shifted" and "hsv" as well and got the same issue. The problem also occurs after saving the image via plt.savefig(). I also tried other image formats like svg but it did not change anything.
As suggested in this answer you can set the image interpolation to "nearest", e.g.,
plt.imshow(data, interpolation="nearest")
See here for a discussion of image antialiasing effects with different interpolation methods.
To keep it short: Is there a way to export plots created with methods like
pcolorfast which basically draw pixels as "real" vector graphics?
I tried to do just that using savefig and saving to a PDF but what would happen is that the plot was actually a vector graphic but the parts drawn by pcolorfast(so basically, what is inside the axes) something like a bitmap. - I checked this using Inkscape.
This resulted in really low resolution plots even though the arrays drawn with pcolorfast where about 3000x4000. I achieved higher resolution by increasing the dpi when exporting, but I'd really appreciate a conversion to a real vector graphic.
Edit: I updated my original code by the piece of code below that should serve to illustrate what exactly I am doing. I tried to involucrate the rasterized tip, but it has had no effect. I still end up with a supersmall PDF-file where the plots are acually raster images (png). I am going to provide you with the data I used and the resulting PDF.
http://www.megafileupload.com/k5ku/test_array1.txt
http://www.megafileupload.com/k5kv/test_array2.txt
http://www.megafileupload.com/k5kw/test.pdf
import numpy as np
import matplotlib.pyplot as plt
arr1= np.loadtxt("test_array1.txt")
arr2= np.loadtxt("test_array2.txt")
fig, (ax1, ax2)=plt.subplots(1, 2)
ax1.set_rasterized(False)
ax1.pcolorfast(arr1)
ax2.pcolorfast(arr2, rasterized=False)
plt.show()
fig.set_rasterized(False)
fig.savefig("test.pdf")
I have an NxN array that I am plotting in Python using matplotlib.pyplot.imshow(). N will be very large and I want my final image to have resolution to match. However, in the code that follows, the image resolution doesn't seem to change with increasing N at all. I think that imshow() (at least how I'm using it) has a fixed minimum pixel size that is larger than that needed to show my NxN array with full resolution.
import numpy as np
import matplotlib.pyplot as plt
from matplotlib import cm
array = np.loadtxt("output.dat",unpack=True)
plt.figsize=(30.0, 30.0)
im = plt.imshow(array,cmap='hot')
plt.colorbar(im)
plt.savefig("mandelbrot.pdf")
As you can see in the code above, I've tried messing with plt.figsize to try and increase resolution but to no avail. I've also tried various output formats (.pdf, .ps, .eps, .png) but these all produced images with lower resolution than I wanted. The .ps, .eps, and .pdf images all looked the exact same.
First, does my problem exist with imshow() or is there some other aspect of my code that needs to be changed to produce higher resolution images?
Second, how do I produce higher resolution images?
plt.figsize() will only change the size of the figure in inches while keeping the default dpi. You can set the resolution of the figure by passing the dpi keyword argument when you save the figure:
fig.savefig('filename.extension', dpi=XXX)
So if you have a figure size of 4x6 and save it with dpi=300 you'll end up with an image with 1200x1800 resolution.
You can also set the default figure size and dpi with matplotlibrc.
Is there a way to show the row and column axes when displaying an image with cv2.imshow()? I am using the python bindings for opencv3.0
Not that I am aware of.
However, since you are using Python you are not constrained to use the rudimentary plotting capabilities of OpenCV HighGUI.
Instead, you can use the much more competent matplotlib library (or any of the other available Python plotting libraries).
To plot an image, including a default axis you do
import matplotlib.pyplot as plt
plt.imshow(image, interpolation='none') # Plot the image, turn off interpolation
plt.show() # Show the image window
I'm not sure I fully understand the question due to lack of info.
However you can use OpenCV's draw line function to draw a line from the example points (10,10) to (10,190), and another from (10,190) to (190,190)
On an example image that is 200X200 pixels in size, this will draw a line down the left hand side of the image, and a line along the bottom. You can then plot numbers or whatever you want along this line at increments of X-pixels.
Drawing text/numbers to an image is similar to drawing a line.
Once you have drawn the image, show with the usual image.imshow().
See OpenCV's drawing documentation here:
http://docs.opencv.org/modules/core/doc/drawing_functions.html
And an example to get you going can be found here:
http://opencvexamples.blogspot.com/2013/10/basic-drawing-examples.html#.VMj-bUesXuM
Hope this helps.
How can I save Python plots at very high quality?
That is, when I keep zooming in on the object saved in a PDF file, why isn't there any blurring?
Also, what would be the best mode to save it in?
png, eps? Or some other? I can't do pdf, because there is a hidden number that happens that mess with Latexmk compilation.
If you are using Matplotlib and are trying to get good figures in a LaTeX document, save as an EPS. Specifically, try something like this after running the commands to plot the image:
plt.savefig('destination_path.eps', format='eps')
I have found that EPS files work best and the dpi parameter is what really makes them look good in a document.
To specify the orientation of the figure before saving, simply call the following before the plt.savefig call, but after creating the plot (assuming you have plotted using an axes with the name ax):
ax.view_init(elev=elevation_angle, azim=azimuthal_angle)
Where elevation_angle is a number (in degrees) specifying the polar angle (down from vertical z axis) and the azimuthal_angle specifies the azimuthal angle (around the z axis).
I find that it is easiest to determine these values by first plotting the image and then rotating it and watching the current values of the angles appear towards the bottom of the window just below the actual plot. Keep in mind that the x, y, z, positions appear by default, but they are replaced with the two angles when you start to click+drag+rotate the image.
Just to add my results, also using Matplotlib.
.eps made all my text bold and removed transparency. .svg gave me high-resolution pictures that actually looked like my graph.
import matplotlib.pyplot as plt
fig, ax = plt.subplots()
# Do the plot code
fig.savefig('myimage.svg', format='svg', dpi=1200)
I used 1200 dpi because a lot of scientific journals require images in 1200 / 600 / 300 dpi, depending on what the image is of. Convert to desired dpi and format in GIMP or Inkscape.
Obviously the dpi doesn't matter since .svg are vector graphics and have "infinite resolution".
You can save to a figure that is 1920x1080 (or 1080p) using:
fig = plt.figure(figsize=(19.20,10.80))
You can also go much higher or lower. The above solutions work well for printing, but these days you want the created image to go into a PNG/JPG or appear in a wide screen format.
Okay, I found spencerlyon2's answer working. However, in case anybody would find himself/herself not knowing what to do with that one line, I had to do it this way:
beingsaved = plt.figure()
# Some scatter plots
plt.scatter(X_1_x, X_1_y)
plt.scatter(X_2_x, X_2_y)
beingsaved.savefig('destination_path.eps', format='eps', dpi=1000)
In case you are working with seaborn plots, instead of Matplotlib, you can save a .png image like this:
Let's suppose you have a matrix object (either Pandas or NumPy), and you want to take a heatmap:
import seaborn as sb
image = sb.heatmap(matrix) # This gets you the heatmap
image.figure.savefig("C:/Your/Path/ ... /your_image.png") # This saves it
This code is compatible with the latest version of Seaborn. Other code around Stack Overflow worked only for previous versions.
Another way I like is this. I set the size of the next image as follows:
plt.subplots(figsize=(15,15))
And then later I plot the output in the console, from which I can copy-paste it where I want. (Since Seaborn is built on top of Matplotlib, there will not be any problem.)