Timeline for Bi-Directional A* function in MATLAB
Current License: CC BY-SA 4.0
17 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Jun 27, 2024 at 19:40 | comment | added | Cris Luengo | "this was replaced with vectorization which has sped up the function" Please post the updated function then. You don't have any answers yet, so updating the code should be fine. It will prevent an answer from demonstrating something you already know how to do. | |
| Jun 27, 2024 at 19:38 | comment | added | Cris Luengo |
Keep the main script short. Just create some bogus data with rand, then call AStarPathTD. Just to demonstrate how your larger program calls this function. MATLAB is not typed, the variable name says nothing about what the variable actually is. So showing some example input values demonstrates what those variables are when you run your larger program.
|
|
| Jun 27, 2024 at 19:32 | comment | added | Cris Luengo |
I'm looking at the code, the script at the end I mainly wanted to see to learn how you call this AStarPathTD, so that I know what each variable is. But the script does not define most of the variables used when calling this function! How is nodes defined? What are Kd, Kt and Ke? Etc. I can't really give advice on speeding up this code if I don't know what these variables are, other than saying "pay attention to the warnings in the MATLAB editor, there are many places where you enlarge arrays in a loop, this is very expensive, the editor recommends you reallocate, do that!"
|
|
| Jun 26, 2024 at 10:09 | history | edited | toolic | CC BY-SA 4.0 |
added 25 characters in body
|
| Jun 26, 2024 at 2:48 | comment | added | LeftFix | The reason that I am not putting in the entire code is because the main script is a couple thousand lines - including all functions | |
| Jun 26, 2024 at 2:46 | history | edited | LeftFix | CC BY-SA 4.0 |
added updated code as well as the helper codes as well as attempting to answer some questioions
|
| Jun 24, 2024 at 18:03 | comment | added | J_H | Show us timing results, please. For a single threaded (one core) workload of interest, show us where profiling says most of the time was spent. | |
| Jun 24, 2024 at 15:19 | comment | added | Cris Luengo | I suggest you post here all the relevant functions, and a “main” script that shows (maybe with bogus data) how the function is called. Then we can run it ourselves. Even if we don’t want to actually run it, being able to do so means we can fully understand the program, nothing is missing. | |
| Jun 24, 2024 at 15:17 | comment | added | Cris Luengo |
When you say “put the code into a cluster” do you mean you are running this function once, and the parfor spreads out the work across clusters? Or are you running this function many times, for different inputs, on the cluster? What is calculateCost? If we don’t know what is happening inside the parfor loop, how are we supposed to understand what the computational cost of the loop is?
|
|
| Jun 24, 2024 at 15:12 | comment | added | Cris Luengo | Please show how you call the function, because if we don’t know what the input arguments are, we can’t really understand what the code does. | |
| Jun 24, 2024 at 15:11 | comment | added | Cris Luengo | Have you profiled the code? | |
| Jun 24, 2024 at 10:34 | history | edited | toolic | CC BY-SA 4.0 |
deleted 8 characters in body; edited title
|
| Jun 24, 2024 at 1:44 | history | edited | LeftFix | CC BY-SA 4.0 |
added 151 characters in body
|
| Jun 24, 2024 at 1:37 | history | edited | Billal BEGUERADJ | CC BY-SA 4.0 |
aerated the text
|
| Jun 24, 2024 at 0:27 | history | edited | LeftFix | CC BY-SA 4.0 |
added 190 characters in body
|
| S Jun 24, 2024 at 0:24 | review | First questions | |||
| Jun 24, 2024 at 1:37 | |||||
| S Jun 24, 2024 at 0:24 | history | asked | LeftFix | CC BY-SA 4.0 |