std::copy a int vector but withing subtraction

[CakeByTheOcean]

I start with a int vector `row` and would like copy it to a address `ptr`.

But now I want to use a sinple way to save the differ with the min_element into the `ptr`. Can I do the job in one line?

Hint: we would get the min using `std::min`

  std::vector<uint64_t> row {16, 32, 30, 40, 50, 60, 70, 80, 90, 16};

  std::copy(row.begin(), row.end(), ptr);  // do the job but no substraction

  // what I want but in one line 
  // firstly calculate the differ 
  std::vector<uint64_t> differ {0, 16, 14, 24, 34, 44, 54, 64, 74, 0};
  // then try to std::copy
    std::copy(differ.begin(), differ.end(), ptr);  // do the job but no substraction

[lastchance]

#include <iostream>
#include <valarray>
using namespace std;

int main()
{
  valarray<uint64_t> row{ 16, 32, 30, 40, 50, 60, 70, 80, 90, 16 };
  valarray<uint64_t> differ = row - row.min();
  
  for ( auto e : row    ) cout << e << " ";   cout << '\n';
  for ( auto e : differ ) cout << e << " ";   cout << '\n';
}

[CakeByTheOcean]

@lastchance
Thanks.
I do my apologize to the statement. I want to combine (the substation and the copy) in one line.
Is that possible?

[lastchance]

valarray<uint64_t> differ = row - row.min();
Doesn't that do it? Call it ptr instead of differ if you want.

I suppose you could also use http://www.cplusplus.com/reference/algorithm/transform/
if you are dead set on vectors.

[CakeByTheOcean]

  // what I want but in one line 
  // firstly calculate the differ 
  std::vector<uint64_t> differ {0, 16, 14, 24, 34, 44, 54, 64, 74, 0};
  // then try to std::copy
    std::copy(differ.begin(), differ.end(), ptr);  // do the job but no subtraction 

After the subtraction, I would like to copy the modifed number to `ptr` this address.
I suppose `std::copy` is 100% what we need. The thing is how to modify the vector inside the `std::copy` function call.

[Grime]

I want to combine (the substation and the copy) in one line.

Are you saying that you don't know how to print the two values in one line?
hint: '\n' is what tells the cursor to go to the next line. 0_0

Anyways if this is a homework don't just simply copy lastchance's code you will definitely get caught.

[CakeByTheOcean]

EDIT:
I want to combine (the substation and the std::copy) in one line.

I think there some misunderstood in my problem statement.
I suppose `std::copy` is 100% what we need. The thing is how to modify the vector inside the `std::copy` function call.

I am not asking with '\n'.

[lastchance]

I'm missing your point, @CakeByTheOcean.

To avoid this XY problem getting even more obscure could you explain the broader picture, please. i.e. give a complete compileable code, and don't try to paraphrase whatever instruction you have been given.

Also, has "ptr" been pre-declared, and if so, what type of variable/container is it?


By the way, a "substation" is a bunch of electrical transformers stepping the power supply voltage up or down. I think you mean "subtraction".

[CakeByTheOcean]

Yes. My bad. I mean substation.
We assume the ptr ist `uint32_t*` with pointing well allocated memory.

[Grime]

You can't do what you're saying with copy. copy (InputIterator first, InputIterator last, OutputIterator result); copy has only one overload and all of its arguments are for inputs and output ONLY.

Does the question need you to write your own function (and hence "one line")? In that case you could have just copy-pasted lastchance's code what's the confusion?

[CakeByTheOcean]

Hmm Not so strict. The reason that I ask, is that I would like to write it more beautiful but have also doubt, if it is possible. I tried with some lambda expression but no success.

[Grime]

Ahh so you thought the third parameter for copy was a functor, makes sense.

[lastchance]

Here you go, @CakeByTheOcean ... all in one line. I prefer my valarray version, though.

#include <iostream>
#include <vector>
#include <algorithm>
using namespace std;

int main()
{
   vector<uint64_t> row{ 16, 32, 30, 40, 50, 60, 70, 80, 90, 16 };
   uint32_t *ptr = new uint32_t[row.size()];
   for ( int i = 0, m = *min_element( row.begin(), row.end() ); i < row.size(); i++ ) ptr[i] = row[i] - m;

   for ( auto e : row    ) cout << e << " ";   cout << '\n';
   for ( int i = 0; i < row.size(); i++ ) cout << ptr[i] << " ";   cout << '\n';
   delete [] ptr;
}

[Grime]

But if you were specified to use std::min then you will need at least two for-loops, one for finding min and another for the actual copying.

Or you would need one really ugly for-loop:
for( int i = 0, min = *row.begin(), bool = 1; i < row.size(); i++) { if(bool) min = std::min(row[i], min) else ptr[i] = row[i]; if(i==row.size() && bool--) i = 0; }

Hey okay maybe the for-loop has a block of statements but it's still one line ;)
And in that way since compilers don't read whitespaces you can fit an entire program worth 5 million lines of code into one line, but what have you gained?