+ 3
Challenge question
- If you had 30 Million x and y values, how would you plot them? - Python only
9 Réponses
+ 2
Drawing huge data on small chart, you could round the numbers to nearest pixel before plot it.
R language are good at plotting big data chart.
+ 4
no i mean read from file or what?
+ 3
30 million x and y stored in what format?
+ 3
[Initial thought before sleep] I'd spray the data at the graphics hardware as close to directly / in parallel as possible, using some lower-level import.
This is what Python's good at anyway (delegating as a 4GL, not trying to be a 3GL).
Reasoning: I used to add assembly direct-to-display-buffer plotting to my graphics codes because standard libraries were too slow.
Now we can leverage the GPU data bus; Python would send a data blob, and then inform the GPU of its structure.
Lest this discourage anyone (because you can't get to that hardware here), you could still emulate the idea on a smaller scale with a square array/buffer you call 'the output'.
Also:Numpy works here, and it's good with rectangular matrices, the shape video buffers come in too.
+ 1
Graphing algorithm Abdur-Rahmaan Janhangeer
+ 1
Store them as arrays. And it is sorta impossible to plot 30 million points unless you have a super computer
+ 1
A file Abdur-Rahmaan Janhangeer
+ 1
Are they integers, or real numbers? What is the organization of the points? Are they sorted in either axis? Do they have multiple Y values per X (or vice-versa)? Do they describe an image, or is the expected result a graph? If a graph, what kind?
Consider sampling the data down to the display resolution.
If there are multiples of the same X and Y points, then you might take a scientific visualization approach and show color or brightness differences as a point is re-used more frequently.