This is with reference to documentation about use of array in D (https://dlang.org/spec/arrays.html#array-setting).

In languages tailor-made for data mining, like R, a subset of an array (not necessarily contiguous values) can be extracted as shown below:

vec //a vector or array in R
c(1:length(vec))[vec > 3] //extract a new array having values of vec that are greater than 3.

I have read about slicing, but it can only subset contiguous values. Correct me if I am wrong here.

How can the above functionality be achieved in D ?
(I am totally new to D)

On Thursday, 14 July 2016 at 10:43:12 UTC, Sahil wrote:
> This is with reference to documentation about use of array in D (https://dlang.org/spec/arrays.html#array-setting).
>
> In languages tailor-made for data mining, like R, a subset of an array (not necessarily contiguous values) can be extracted as shown below:
>
> vec //a vector or array in R
> c(1:length(vec))[vec > 3] //extract a new array having values of vec that are greater than 3.
>
> I have read about slicing, but it can only subset contiguous values. Correct me if I am wrong here.
>
> How can the above functionality be achieved in D ?
> (I am totally new to D)

easily. The function you are looking for is called 'filter'

import std.algorithm;
import std.array;
int[] a = [ 4,5,8,1,3,2,9,10];
auto b = a.filter!(e => e>3).array; //b contains the elements 4,5,8,9,10

On Thursday, 14 July 2016 at 10:51:33 UTC, Nicholas Wilson wrote:
> On Thursday, 14 July 2016 at 10:43:12 UTC, Sahil wrote:

>> (I am totally new to D)
>
> easily. The function you are looking for is called 'filter'
>
> import std.algorithm;
> import std.array;
> int[] a = [ 4,5,8,1,3,2,9,10];
> auto b = a.filter!(e => e>3).array; //b contains the elements 4,5,8,9,10

I should explain this a bit more if you are totally new to D.

although filter looks like a method for an array type it is not it is a function.
in D function like f(a,b) can be rewritten as a.f(b);

the '!' is the template instansiation operator (as filter is a template function) and its template parameter is the predicate to filter in this case ,(e => e>3).

As with anyone totally new to D i cannot recommend more highly Ali's book.

On Thursday, 14 July 2016 at 10:51:33 UTC, Nicholas Wilson wrote:
> On Thursday, 14 July 2016 at 10:43:12 UTC, Sahil wrote:
>> This is with reference to documentation about use of array in D (https://dlang.org/spec/arrays.html#array-setting).

> easily. The function you are looking for is called 'filter'
>
> import std.algorithm;
> import std.array;
> int[] a = [ 4,5,8,1,3,2,9,10];
> auto b = a.filter!(e => e>3).array; //b contains the elements 4,5,8,9,10

Thanks for a prompt reply. This gives me confidence to explore D for faster implementations in data mining.
Thanks a lot :-)

On Thursday, 14 July 2016 at 11:01:20 UTC, Nicholas Wilson wrote:
> On Thursday, 14 July 2016 at 10:51:33 UTC, Nicholas Wilson wrote:

> I should explain this a bit more if you are totally new to D.
>
> although filter looks like a method for an array type it is not it is a function.
> in D function like f(a,b) can be rewritten as a.f(b);
>
> the '!' is the template instansiation operator (as filter is a template function) and its template parameter is the predicate to filter in this case ,(e => e>3).
>
> As with anyone totally new to D i cannot recommend more highly Ali's book.

Yes, I have started with this book.
Thanks for the explanation!

On 7/14/16 7:03 AM, Sahil wrote:
> On Thursday, 14 July 2016 at 10:51:33 UTC, Nicholas Wilson wrote:
>> On Thursday, 14 July 2016 at 10:43:12 UTC, Sahil wrote:
>>> This is with reference to documentation about use of array in D
>>> (https://dlang.org/spec/arrays.html#array-setting).
>
>> easily. The function you are looking for is called 'filter'
>>
>> import std.algorithm;
>> import std.array;
>> int[] a = [ 4,5,8,1,3,2,9,10];
>> auto b = a.filter!(e => e>3).array; //b contains the elements 4,5,8,9,10
>
> Thanks for a prompt reply. This gives me confidence to explore D for
> faster implementations in data mining.
> Thanks a lot :-)

Please be aware, however, that filter is going to still iterate the array. So the subset operation may not be as performant as a subset operation and data type that is written to deal with these kinds of operations.

-Steve