cubemovie

Python code for transmuting numpy arrays into reasonable mp4 movies using matplotlib and ffmpeg.

Basic usage

Hold numpy array in a .npy file.

Generate a .npy file from numpy.save(filename.npy,array).

A preprocessing script can produce all the frames and save them to a 3-D array of shape:

[N_frames,N_vertical,N_horizontal] 

Then, run:

python -m cubemovie.basic filename.npy

Flags

-l # uses lognorm to scale colors instead of linear
-s # show movie instead of saving
-m # alternate mpeg option for wider compatibility but worse quality
-g # make a gif from the mp4 output
-h # show help
-fps N # set fps 
-size INCHES # set matplotlib figsize in inches (higher will be higher res)

Advanced usage example:

python -m cubemovie.basic filename1.npy -l
python -m cubemovie.mask  filename1.npy ignored_filename2.npy -l -s
python -m cubemovie.basic filename1.npy -fps 5

Files

parse.py parses the command line arguments: flags and filenames

tools.py contains some useful functions

basic.py is the main script

mask.py is a specific application

dict.py is a specific application

Mask

Given an array of shape [2,N_frames,N_vertical,N_horizontal] split into two arrays of shape [N_frames,N_vertical,N_horizontal] The first array is some original field, and the second array is like a mask: non-zero values in some areas, and 0 everywhere else. Both arrays are plotted side-by-side, and contours of the Mask (second array) are drawn over the Original (first array).

Dict

Given a python dictionary, plots the following according to the dictionary key:

contour-1: takes a mask of 0 and 1 and makes 2-D contour plot to delineate the boundaries in black

contour-2: takes a mask of 0 and 1 and makes 2-D contour plot to delineate the boundaries in red

scatter-x, scatter-y: given these, draw scatter points on top

data: