Persistent drawings with matplotlib animations - python

I am trying to implement Alec Radford's visualizations of gradient descents in Python with matplotlib animation. I have found this question, which is a good start and I have been playing around a bit with the code.
However I am wondering how to animate persistent lines.
(For example the ones that show the different path taken by the algorithm for different gradient descent in Alec's visualizations).
Maybe a link to the code Alec used is available it certainly looks like it was also done in matplotlib...

Related

Simple 3D Graphics in Python

I am working in a project where 3D visualizations are important to see what is happening during the setup stage and perhaps for visual validation by making a short videos of what is happening.
The problem that I have is that 3D visualizations in Python are too sophisticated, and complicated to learn for what I need. I find that Mathematica is the perfect kind of software...but it is not portable and is very expensive.
Question
Is there any Python package similar to Mathematica?
Clarification
I don't want a "plotting" program, since plotting is not what I am looking for. I want to generate simple geometric shapes like spheres and cubes that can move around, this is more than enough. Give some coordinates, perhaps a rotation, and the program just shows the desired image(s) to export as a .png or make a quick video; as in Mathematica.
Packages like Pygame, Panda3D, Pyglet, etc., look too complicated and an overkill for what I need, as well as software like Blender, etc. Jupyter notebooks are similar, but they don't have the 3D graphics capabilities. I found a Python module named Fresnel, but it looks too sophisticated for what I need.
I have read several answers to this question here in Stack Overflow, but they seem outdated and not really what I am looking for.
Further Clarification
To draw spheres in Mathematica you do:
coordinates = Flatten[Table[Table[Table[ {i,j,k}, {k,1,10}], {j,1,10}], {i,1,10}],1]
spheres= Flatten[Table[Graphics3D[{Sphere[coordinates[[i]],0.5]}],{i,1,Length[coordinates]}]]
Show[{spheres}]
This is a simple quick and easy way of displaying a group of spheres. To use any program in Python to do the same, it seems like you must be an expert in 3D graphics to do this simple thing.
Programs that have capabilities of using Python scripts, like Blender, make it difficult to use the interface in a straight forward way (try doing the same in Blender, it will take a while just to learn the basics!).
I know several other user-friendly plotting libraries than matplotlib, but not a lot provide an interactive view. There is of course the well known vtk but it's not for end-user
plotly
For usage in a notebook, like jupyter and mathematica, you probably would go for plotly It's using a browser-based interface with plots very similar to mathematica
pymadcad
If you need a more offline version and what you are looking for is some view you can rotate/zoom/pan to look on your geometry by different sides, you can take a look at pymadcad
It even works with touchscreens.
It is not centered on 3D visualization, so it's a bit overkill to use it only for it, but for 3D curves, 3D surfaces, spheres and cubes as you said, it can do the job
simple plots with pymadcad:
from madcad import *
from madcad.rendering import Displayable
from madcad.displays import GridDisplay
# create a wire from custom points
s = 100
mycurve = Wire([ vec3(sin(t/s), cos(t/s), 0.1*t/s)
for t in range(int(s*6*pi)) ])
# create a sphere
mysphere = uvsphere(vec3(0), 0.5)
# display in a separated window
show([
mycurve, # displaying the curve
mysphere, # displaying the sphere
Displayable(GridDisplay, vec3(0)), # this is to have a 2D grid centered on origin
])
result:
(The window is dark because so is my system theme, but likely it will adapt to yours)

Simplest way to publish matplotlib-like interactive animation to html?

I'm using matplotlib and python to make an animated scatter plot with points as 'balls' that bounce around, just like this:
https://jakevdp.github.io/downloads/videos/particle_box.mp4
It is interactive at runtime (with sliders to change velocity, attraction, etc), simulating on the go.
I would like to then publish it with the interactivity and all in html.
Problem: I don't know of a way to publish matplotlib interactivity to html directly.
So can I do it in python? Or is there a better way?
Which library (or program, if different from python) would you recommend as the simplest and fastest for this kind of project? I don't need "pretty" customizations and all that, I just need easy calculations and simulations (python) to then plot in a few easy lines of code (matplotlib). I do need to change the dots (balls) colors though.
I have looked at mpld3 that should wrap matplotlib around D3js. But I don't know anything about js and it is not very straightforward, so it would take me a bit to learn.
I have also looked at plotly, but it doesn't seem to have the same customization of the animation I need.
I have looked at Dash with plotly, but it would be a whole new environment to learn, and definitely overkill.
So the questions are: is there a way to output matplotlib interactive animations to html that I haven't found?
If not, what tools would you use to accomplish this project within a fast timeframe and shallow learning curve, based on my beginner/low intermediate python and matplotlib skill level?
Thank you!

Is there an equivalent to the Matlab figure window in Python (with all tools)?

I'm just wondering if it exists an equivalent to the Matlab figure window in Python where we can modify plots directly from the figure window, or add some features (text, box , arrow, and so on), or make curve fitting, etc.
Matplotlib is good, but it is not as high-level as the Matlab figure. We need to code everything and if we want to modify plots, we need to modify the code directly (except for some basic stuffs like modifing the line color)
With matplotlib, you will indeed remain in the "code it all" workflow. This is not directly the answer you expect but the matplotlib documentation recently gained a very instructive figure that will probably help you if you stay with matplotlib: http://matplotlib.org/examples/showcase/anatomy.html shows the "anatomy" of the figure with all the proper designations for the parts of the figure.
Overall, I could always find examples of what I needed in their excellent gallery http://matplotlib.org/gallery.html
In my opinion, you'll save time by coding these customizations instead of doing them by hand. You may indeed feel otherwise but if not there is a ton of examples of matplotlib code on SO, in their docs and a large community of people around it :-)

Edges with Direction in pyqtgraph GraphItem

I am trying to visualize a Control Flow Graph in Python using pyqtgraph. I have the following two problems.
How can I visualize the edges with a direction?
How can I visualize a self edge?
I tried looking into the documentation, but couldn't find. Obviously, I didn't get time to read it all!
While pyqtgraph is awesome, for my use case I found a much better tool to do this.
graphviz is a nice tool to develop Control Flow Graphs quite conveniently, and has a large number of features for this particular problem.
For direction, you might add a pg.ArrowItem at the end of each line (although this could have poor performance for large networks), and for self connections, QtGui.QGraphicsEllipseItem combined with an arrow.

Drawing clustered graphs in Python

I already have a way of clustering my graph, so the process of clustering isn't the issue here. What I want to do is, once we have all the nodes clustered - to draw the clustered graph in Python, something like this:
I looked into networkx, igraph and graph-tool, but they seem to do the clustering, but not the drawing. Any ideas and propositions of what library should I use for drawing the already clustered graph, which will minimize the number of crossing links?
Take a look at GraphViz
http://www.graphviz.org/Gallery/directed/cluster.html
There's a Python binding for that, but I have to say I always create the text files directly as they're easy enough to write. Don't be fooled by the plain-looking examples, every aspect of your graph is highly customizable and you can make some pretty nifty graph visualizations with it. Not sure about nested clusters though, never tried that out.

Categories