Use Property Node

[altenbach]

---

LSASS wrote:

I am usind the Array subset to split my huge 1D array in to chunks of 8. But I am unsure as to what to pass in to the index input of the Array subset function. When I put it in a for loop, the index continuously changes. So what do I input?

---

Split it into what? Please show us your code so we have a better idea.

The index is zero at the first interation and increases by one for each later iteration. Multiply it by 8 and wire to the index input and set the size to 8 and you get the next eight elements per iteration.

As has been mentioned before, "reshape array" is probably the better tool.

Message Edited by altenbach on 05-10-2009 03:43 PM

[smercurio_fc]

---

LSASS wrote:

Hello smercurio,

I am usind the Array subset to split my huge 1D array in to chunks of 8. But I am unsure as to what to pass in to the index input of the Array subset function. When I put it in a for loop, the index continuously changes. So what do I input?

---

You input the next multiple of 8, based on the current iteration value. When iteration = 0, the value you input is 0. When iteration = 1, you input 8. When iteration = 2, you input 16. When ...

The Reshape Array method would not require this since the for-loop can be auto-indexed from the resultant 2D array. 

Message Edited by smercurio_fc on 05-10-2009 09:05 PM

[9094342278]

Hi

   I think use  property node,i am not conform , try it.........

Regards

Munna

[tinnitus]

hello lsass,

i don't know if your sequence can be present more than  one time or if your data are double or not or are 

exactly the same.

Here is a suggestion 

Regards

Tinnitus

[Norbert_B]

---

LSASS wrote:

[...]I have a problem: I have a huge 1D array and I need to split it in to sets of 8 elements as need to compare these 8 elements with another  1 D array of size 8. [...]

---

Well, coming a little late into this discussion...but reading this sentence sounds like a pattern match to me. Maybe this is better in regard to performance than to split the array and do the comparison afterwords.....

just my 5 cents,

Norbert 

[James_R]

Smercurio

Great Idea for decimation

LSASS is what you are looking for?

[altenbach]

James, you can eliminate the "AND array elements" by setting the comparison mode to "compare aggregates".

[James_R]

Thanks !  that is a cool feature

[tinnitus]

hello,

you could also swap  0 and 1 reshape entries and replace number to decimate command with

size of sequence looked (3) , it will avoid to use a transpose array function.

regards

Tinnitus

[James_R]

Tinnitus  I have to dissagree with your method to eliminate the transpose array

In this example the number of "Channels"  = the number that you wish to decimate by which = 4

and there are 3 samples for each channel

applying your proposed changes does not yield the same dataset   (you can not simply change the number to decimate from 4 to 3) because the array size of the match pattern is how many samples not channels

the only way i can think of to eliminate the transpose array is to index the array manually instead of using the autoindex