Daily Coding Problem #2

Seeing the first post gained some popularity, here is the second problem

For quite some time I found my time procastinating after getting home from work. Untill a few days ago I stumbled upon the Daily Coding Problem (DCP) and decided to give it a shot.
The code is in C#.

How it works?

The folks at DCP send you a problem that was asked at a top company everyday for you to solve. The premium membership also offers the opportunity to verify your solution.

Problem #2

This problem was asked by Uber.

Given an array of integers, return a new array such that each element at index i of the new array is the product of all the numbers in the original array except the one at i.

For example, if our input was [1, 2, 3, 4, 5], the expected output would be [120, 60, 40, 30, 24]. If our input was [3, 2, 1], the expected output would be [2, 3, 6].

Follow-up: what if you can't use division?

My solution

using System;
using System.Linq;

namespace Task02
{
    public class Program
    {
        public static void Main(string[] args)
        {
            var input = Console.ReadLine()
                .Split(' ')
                .Select(int.Parse)
                .ToArray();

            var result = new int[input.Length];

            for (int i = 0; i < result.Length; i++)
            {
                result[i] = 1;

                for (int j = 0; j < result.Length; j++)
                {
                    if (j != i)
                    {
                        result[i] *= input[j];
                    }
                }
            }


            Console.WriteLine(string.Join(' ', result));
        }
    }
}

Explanation

Here I could not think of any better than the naive solution, which works in O(n^2) time and you never like square complexity.
Basically for every item in the result array, foreach the input array and check if the index is different from the item's. If they are, multiply the current value of the item times the current input item

I will try to do the daily problem everyday, but sometimes life gets in the way :)
Solutions are posted in my github account

cwetanow / DailyCodingProblem

DailyCodingProblem

This repository contains solutions of the Daily Coding Problem tasks

View on GitHub

.

Feel free to leave a like and let me know in the comments if I should continue to post my solutions.

If you also have any better solution in mind, by all means share it, so we can learn from each other.

Comments

[edh_developer]

In python 2.7, no division, O(n) time. The tradeoff is using a bunch of memory, if the list is very long.

values = map(int,raw_input().split(' '))


def getLeftProducts(rawValues):
        leftProducts = []

        if len(rawValues) > 0:
                total = 1
                for val in rawValues:
                        total *= val
                        leftProducts.append(total)

        return leftProducts


def getRightProducts(rawValues):
        rightProducts = []

        if len(rawValues) > 0:
                total = 1
                index = len(rawValues) - 1

                while index >= 0:
                        total *= rawValues[index]
                        rightProducts.insert(0,total)
                        index -= 1

        return rightProducts


def processList(rawValues):
        if len(rawValues) < 2:
                return rawValues

        resultList = []
        leftList = getLeftProducts(rawValues)
        rightList = getRightProducts(rawValues)

        resultList.append(rightList[1])

        index = 1
        while index < (len(rawValues) - 1):
                resultList.append( leftList[index - 1] * rightList[index + 1] )
                index += 1

        resultList.append(leftList[index - 1])

        return resultList


print ' '.join(map(str,processList(values)))

[Albin Severinson]

Nice find :-)

For the first problem I'd loop over the array twice. The first time to compute the total product and the second time to fill in the output array, dividing out each element of the input array from the total product for the corresponding output element.

I'm guessing that your solution is for the follow-up question though :-). If you have logarithms you could use them to convert multiplication and division to addition and subtraction. This is a common trick when working over finite fields.

[Benjamin Braatz]

Or with no additional space required:

def prod(input):
    output = []
    right = 1
    for number in reversed(input):
        output.insert(0, right)
        right *= number
    left = 1
    for i, number in enumerate(input):
        output[i] *= left
        left *= number
    return output

[Stefano Prosperi]

O(n) with division, in C#

using System;
using System.Collections.Generic;
using System.Linq;

namespace Daily_Coding_Problem_2
{
    public class Program
    {
        static void Main(string[] args)
        {
            var numbers = GetNumberList();

            var ret = Problem2(numbers);

            foreach (var i in ret)
            {
                Console.WriteLine(i);
            }
        }

        public static IEnumerable<int> GetNumberList()
        {
            return Console.ReadLine() ?
                              .Split(' ')
                              .Select(int.Parse)
                              .ToArray();
        }

        public static IEnumerable<int> Problem2(IEnumerable<int> numbers)
        {
            var enumerable = numbers as int[] ?? numbers.ToArray();

            var product = enumerable.Aggregate(1, (current, n) => current * n);

            return enumerable.Select(n => product / n).ToList();
        }
    }
}

[Luiz Guilherme]

stackblitz.com/edit/js-xqagwj

const appDiv = document.getElementById('app');

const arr = [1, 2, 3, 4, 5];

const result = arr.map(
  (item) => 
    arr
      .filter(n => n !== item )
      .reduce((acc,current) => acc * current , 1)
);


appDiv.innerHTML = `<h1>RESULT: ${result.toString()}</h1>`;

[Hasheen DeBerry]

I did something very similar to this. I realised quickly that if any of the numbers in the array are repeated, the totals don’t come out right. I had to explicitly remove the element from the input array and then use the reduce function to calculate the product.
HTH

[Gungrave223]

const array = [1, 2, 3, 4, 5];

const updated = array.map((ignored, currentIndex) =>
  array.reduce((accum, value, index) => {
    if (currentIndex === index) return accum
    return accum * value;
  }, 1));

[Benjamin Braatz]

Solution in Python without division, still in O(n²), but inner loop is only executed n²/2 times:

def prod(input):
    output = []
    product = 1
    for number in input:
        for i in range(len(output)):
            output[i] *= number
        output.append(product)
        product *= number
    return output

[Markus]

Here's my TypeScript version.

This approach works in O(n) by dividing the product of all elements by x[i].

function uberMultiply(x: Array<number>) {
  const product = x.reduce( (a,b) => a * b );
  return x.map(v => product / v);
}

[Kasey Speakman]

Without being aware of possible constraints on the input numbers, a solution with division is most efficient as it conserves both memory (2 array allocations: input and output) and cpu (2 enumerations of the the array, O(2n) ~= O(n)). This was described by @severinson . Here it is in F#.

let calculate arr =
    let product = arr |> Array.reduce (*)
    arr |> Array.map (fun x -> product / x)

To consider a solution without division, understand that we are using division here to reverse a multiplication. An alternative is to choose not to multiply in the first place. But in order to do that, we move to an O( n² ) solution where we shape each element of the output array as we process the input. Then we can choose to selectively skip multiplication.

let calculate arr =
    // initialize the output array to 1s in prep for multiplication
    let output = Array.init (Array.length arr) (fun _ -> 1)
    arr |> Array.iteri (fun i x ->
        output |> Array.iteri (fun j y ->
            if i <> j then output.[j] <- y * x
        )
    )
    output

Note: This function is "pure", despite using mutation internally.

In some branches of mathematics certain operations are not reversible, such as matrix cross product in linear algebra. Perhaps better knowledge of that discipline would yield more efficient algorithms. However, all definitions of this sequence in OEIS use division for calculations.

[mdusoemsa]

Part 2 solution, slightly better than O(n^2), using a copy to initialize the array, and the inner loop runs 2 fewer times than the outer.

namespace Playground
{
    class Program
    {
        static void Main(string[] args)
        {
            var input = new int[] {1, 2, 3, 4, 5};

            var output = ComputeProducts(input);

            foreach (int i in output)
            {
                Console.Write($"{i}\t");
            }

            Console.ReadLine();
        }

        static int[] ComputeProducts(int[] source)
        {
            var count = source.Length;
            // making a few assumptions for edge cases...
            if (count < 2)
                return source;

            if (count == 2)
                return new[] {source[1], source[0]};

            var result = new int[count];
            // initialize the array, eliminating the need to do a "set to 1" step
            Array.Copy(source,1,result,0,count - 1);
            result[count - 1] = source[0];

            // loop the result
            for (int i = 0; i < count; i++)
            {
                // we skip the current index, and the next one has already been set above
                for (int j = 2; j < count; j++)
                {
                    result[i] *= source[(j + i) % count];
                }
            }

            return result;
        }
    }
}

[Ram Kumar R S]

def remaining_product(input_array:list)->list:
    new_array = [1 for i in range(len(input_array))]  # output
    for i in range(len(input_array)):
        right_index=[j for j in range(i+1,len(input_array))]
        left_index = [j for j in range(i)]
        all_index= right_index + left_index
        for index in all_index:
            new_array[i] *= input_array[index]
    return (new_array)