24 lines
1.3 KiB
Markdown
24 lines
1.3 KiB
Markdown
|
Now, after two pseudo-pathfinding problems we finally get a real one.
|
||
|
|
||
|
The right solution would proably be to search online for dijkstras algorithm,
|
||
|
but I'm lazy and I kinda remember the essence - so I implement my own algorithm from what I remember about pathfinding :D.
|
||
|
|
||
|
The general idea is still the same - we generate a 80x80 grid where each cell contains the minimal distance to the node `[0,0]`.
|
||
|
At the end we just have to look at the value of `distance[79, 79]`.
|
||
|
Initially all distances are set to an absurdly high number.
|
||
|
|
||
|
Additionally we have a a 80x80 marker-grid where we mark cells either `UNKNOWN (#)`, `MARKED (0)` or `FINISHED (1)`.
|
||
|
Initially all cells are Unknown.
|
||
|
|
||
|
We start our algorithm by setting `distance[0, 0] = data[0, 0]` and `marker[0, 0] = MARKED`
|
||
|
|
||
|
Then we execute the main loop until all cells are marked with FINISHED,
|
||
|
in our main loop we do:
|
||
|
|
||
|
- Search fo a MARKED cell
|
||
|
- Calculate for all 4 directions the distance:
|
||
|
- `distance = distance[x, y] + data[x+1, y]` *(or `x-1`, `y+1`, `y-1`, depending on the direction)*
|
||
|
- If the calculated distance is less than the cached value (`distance[x+1, y]`) update the distance and set the updated cell to MARKED
|
||
|
- Set our current cell to FINISHED (`marker[x, y] = FINISHED`)
|
||
|
|
||
|
Rinse and repeat until finished.
|