Create a two-dimensional array at runtime: Difference between revisions
Create a two-dimensional array at runtime (view source)
Revision as of 20:08, 11 May 2011
, 13 years ago→{{header|Go}}: more techniques
(→{{header|Go}}: added note on array composition) |
(→{{header|Go}}: more techniques) |
||
Line 427:
printfn "%d" arr.[0,0]</lang>
=={{header|Go}}==
Arrays in Go are only one dimensional. Code below show the obvious way of composing a 2d array as an array of arrays that can be indexed as a[r][c]
<lang go>package main
Line 455:
a = nil
// memory allocated earlier with make can now be garbage collected.
}</lang>
The technique above alocates each row separately. This might be good if you need extremely large arrays that cannot be allocated in a single piece. It might be bad though, for locality, as there would be no guarantee that the separate allocations would be localized in memory. A technique that maintains locality is this,
<lang go> // allocate composed 2d array
a := make([][]int, row)
e := make([]int, row * col)
for i := range a {
a[i] = e[i*col:(i+1)*col]
}</lang>
Now all rows are allocated with a single allocation. Alternatively, slice e can be used directly without going through slice a. Element r c can be accessed simply as e[r*cols+c] for example, or accessor functions can be defined such as,
<lang go>
func get(r, c int) int {
return e[r*cols+c]
}</lang>
| |||