área rectangular más grande en un histograma | Serie 1

Encuentre el área rectangular más grande posible en un histograma dado donde el rectángulo más grande puede estar formado por varias barras contiguas. Para simplificar, suponga que todas las barras tienen el mismo ancho y el ancho es 1 unidad. 

Por ejemplo, considere el siguiente histograma con 7 barras de alturas {6, 2, 5, 4, 5, 1, 6}. El rectángulo más grande posible es 12 (vea la figura a continuación, el rectángulo de área máxima está resaltado en rojo)
 

histogram

Una solución simple es considerar todas las barras una por una como puntos de partida y calcular el área de todos los rectángulos comenzando con cada barra. Finalmente devuelva el máximo de todas las áreas posibles. La complejidad temporal de esta solución sería O(n^2).

Podemos usar Divide and Conquer para resolver esto en tiempo O(nLogn). La idea es encontrar el valor mínimo en la array dada. Una vez que tenemos el índice del valor mínimo, el área máxima es el máximo de los siguientes tres valores. 

  1. Área máxima en el lado izquierdo del valor mínimo (sin incluir el valor mínimo) 
  2. Área máxima en el lado derecho del valor mínimo (sin incluir el valor mínimo) 
  3. Número de barras multiplicado por el valor mínimo. 

Las áreas a la izquierda y a la derecha de la barra de valor mínimo se pueden calcular recursivamente. Si usamos la búsqueda lineal para encontrar el valor mínimo, entonces la complejidad de tiempo del peor de los casos de este algoritmo se convierte en O (n ^ 2). En el peor de los casos, siempre tenemos (n-1) elementos en un lado y 0 elementos en el otro lado y si el mínimo de búsqueda toma O (n) tiempo, obtenemos una recurrencia similar al peor caso de Quick Sort. 
¿Cómo encontrar el mínimo de manera eficiente? Para esto, se puede usar la consulta de rango mínimo usando el árbol de segmentos . Construimos un árbol de segmentos de las alturas de histograma dadas. Una vez que se crea el árbol de segmentos, todas las consultas mínimas de rango toman tiempo O (Inicio de sesión) . Entonces, sobre toda la complejidad del algoritmo se vuelve.

Tiempo general = Tiempo para construir el árbol de segmentos + Tiempo para encontrar recursivamente el área máxima
El tiempo para construir el árbol de segmentos es O(n) . Sea T(n) el tiempo para encontrar recursivamente el área máxima. Se puede escribir de la siguiente manera. 
T(n) = O(Logn) + T(n-1) 
La solución de la recurrencia anterior es O(nLogn). Entonces, el tiempo total es O (n) + O (nLogn), que es O (nLogn).
A continuación se muestra la implementación en C++ del algoritmo anterior. 

Implementación:

C++

// A Divide and Conquer Program to find maximum rectangular area in a histogram
#include <bits/stdc++.h>
using namespace std;
  
// A utility function to find minimum of three integers
int max(int x, int y, int z)
{  return max(max(x, y), z); }
  
// A utility function to get minimum of two numbers in hist[]
int minVal(int *hist, int i, int j)
{
    if (i == -1) return j;
    if (j == -1) return i;
    return (hist[i] < hist[j])? i : j;
}
  
// A utility function to get the middle index from corner indexes.
int getMid(int s, int e)
{   return s + (e -s)/2; }
  
/*  A recursive function to get the index of minimum value in a given range of
    indexes. The following are parameters for this function.
  
    hist   --> Input array for which segment tree is built
    st    --> Pointer to segment tree
    index --> Index of current node in the segment tree. Initially 0 is
             passed as root is always at index 0
    ss & se  --> Starting and ending indexes of the segment represented by
                 current node, i.e., st[index]
    qs & qe  --> Starting and ending indexes of query range */
int RMQUtil(int *hist, int *st, int ss, int se, int qs, int qe, int index)
{
    // If segment of this node is a part of given range, then return the
    // min of the segment
    if (qs <= ss && qe >= se)
        return st[index];
  
    // If segment of this node is outside the given range
    if (se < qs || ss > qe)
        return -1;
  
    // If a part of this segment overlaps with the given range
    int mid = getMid(ss, se);
    return minVal(hist, RMQUtil(hist, st, ss, mid, qs, qe, 2*index+1),
                  RMQUtil(hist, st, mid+1, se, qs, qe, 2*index+2));
}
  
// Return index of minimum element in range from index qs (query start) to
// qe (query end).  It mainly uses RMQUtil()
int RMQ(int *hist, int *st, int n, int qs, int qe)
{
    // Check for erroneous input values
    if (qs < 0 || qe > n-1 || qs > qe)
    {
        cout << "Invalid Input";
        return -1;
    }
  
    return RMQUtil(hist, st, 0, n-1, qs, qe, 0);
}
  
// A recursive function that constructs Segment Tree for hist[ss..se].
// si is index of current node in segment tree st
int constructSTUtil(int hist[], int ss, int se, int *st, int si)
{
    // If there is one element in array, store it in current node of
    // segment tree and return
    if (ss == se)
       return (st[si] = ss);
  
    // If there are more than one elements, then recur for left and
    // right subtrees and store the minimum of two values in this node
    int mid = getMid(ss, se);
    st[si] =  minVal(hist, constructSTUtil(hist, ss, mid, st, si*2+1),
                     constructSTUtil(hist, mid+1, se, st, si*2+2));
    return st[si];
}
  
/* Function to construct segment tree from given array. This function
   allocates memory for segment tree and calls constructSTUtil() to
   fill the allocated memory */
int *constructST(int hist[], int n)
{
    // Allocate memory for segment tree
    int x = (int)(ceil(log2(n))); //Height of segment tree
    int max_size = 2*(int)pow(2, x) - 1; //Maximum size of segment tree
    int *st = new int[max_size];
  
    // Fill the allocated memory st
    constructSTUtil(hist, 0, n-1, st, 0);
  
    // Return the constructed segment tree
    return st;
}
  
// A recursive function to find the maximum rectangular area.
// It uses segment tree 'st' to find the minimum value in hist[l..r]
int getMaxAreaRec(int *hist, int *st, int n, int l, int r)
{
    // Base cases
    if (l > r)  return INT_MIN;
    if (l == r)  return hist[l];
  
    // Find index of the minimum value in given range
    // This takes O(Logn)time
    int m = RMQ(hist, st, n, l, r);
  
    /* Return maximum of following three possible cases
       a) Maximum area in Left of min value (not including the min)
       a) Maximum area in right of min value (not including the min)
       c) Maximum area including min */
    return max(getMaxAreaRec(hist, st, n, l, m-1),
               getMaxAreaRec(hist, st, n, m+1, r),
               (r-l+1)*(hist[m]) );
}
  
// The main function to find max area
int getMaxArea(int hist[], int n)
{
    // Build segment tree from given array. This takes
    // O(n) time
    int *st = constructST(hist, n);
  
    // Use recursive utility function to find the
    // maximum area
    return getMaxAreaRec(hist, st, n, 0, n-1);
}
  
// Driver program to test above functions
int main()
{
    int hist[] =  {6, 1, 5, 4, 5, 2, 6};
    int n = sizeof(hist)/sizeof(hist[0]);
    cout << "Maximum area is " << getMaxArea(hist, n);
    return 0;
}

Java

// A Divide and Conquer Program to find maximum rectangular area in a histogram
import java.util.*;
  
class GFG{
  static int[] hist;
  static int[] st;
    
  // A utility function to find minimum of three integers
  static int max(int x, int y, int z)
  {  return Math.max(Math.max(x, y), z); }
  
  // A utility function to get minimum of two numbers in hist[]
  static int minVal(int i, int j)
  {
    if (i == -1) return j;
    if (j == -1) return i;
    return (hist[i] < hist[j])? i : j;
  }
  
  // A utility function to get the middle index from corner indexes.
  static int getMid(int s, int e)
  {   return s + (e -s)/2; }
  
  /*  A recursive function to get the index of minimum value in a given range of
    indexes. The following are parameters for this function.
  
    hist   -. Input array for which segment tree is built
    st    -. Pointer to segment tree
    index -. Index of current node in the segment tree. Initially 0 is
             passed as root is always at index 0
    ss & se  -. Starting and ending indexes of the segment represented by
                 current node, i.e., st[index]
    qs & qe  -. Starting and ending indexes of query range */
  static int RMQUtil( int ss, int se, int qs, int qe, int index)
  {
    // If segment of this node is a part of given range, then return the
    // min of the segment
    if (qs <= ss && qe >= se)
      return st[index];
  
    // If segment of this node is outside the given range
    if (se < qs || ss > qe)
      return -1;
  
    // If a part of this segment overlaps with the given range
    int mid = getMid(ss, se);
    return minVal( RMQUtil(ss, mid, qs, qe, 2*index+1),
                  RMQUtil( mid+1, se, qs, qe, 2*index+2));
  }
  
  // Return index of minimum element in range from index qs (query start) to
  // qe (query end).  It mainly uses RMQUtil()
  static int RMQ(  int n, int qs, int qe)
  {
    // Check for erroneous input values
    if (qs < 0 || qe > n-1 || qs > qe)
    {
      System.out.print("Invalid Input");
      return -1;
    }
  
    return RMQUtil(  0, n-1, qs, qe, 0);
  }
  
  // A recursive function that constructs Segment Tree for hist[ss..se].
  // si is index of current node in segment tree st
  static int constructSTUtil(int ss, int se, int si)
  {
    // If there is one element in array, store it in current node of
    // segment tree and return
    if (ss == se)
      return (st[si] = ss);
  
    // If there are more than one elements, then recur for left and
    // right subtrees and store the minimum of two values in this node
    int mid = getMid(ss, se);
    st[si] =  minVal( constructSTUtil( ss, mid,  si*2+1),
                     constructSTUtil( mid+1, se,  si*2+2));
    return st[si];
  }
  
  /* Function to construct segment tree from given array. This function
   allocates memory for segment tree and calls constructSTUtil() to
   fill the allocated memory */
  static void constructST(int n)
  {
    // Allocate memory for segment tree
    int x = (int)(Math.ceil(Math.log(n))); //Height of segment tree
    int max_size = 2*(int)Math.pow(2, x) - 1; //Maximum size of segment tree
    st = new int[max_size*2];
  
    // Fill the allocated memory st
    constructSTUtil( 0, n-1, 0);
  
    // Return the constructed segment tree
    // return st;
  }
  
  // A recursive function to find the maximum rectangular area.
  // It uses segment tree 'st' to find the minimum value in hist[l..r]
  static int getMaxAreaRec( int n, int l, int r)
  {
    // Base cases
    if (l > r)  return Integer.MIN_VALUE;
    if (l == r)  return hist[l];
  
    // Find index of the minimum value in given range
    // This takes O(Logn)time
    int m = RMQ( n, l, r);
  
    /* Return maximum of following three possible cases
       a) Maximum area in Left of min value (not including the min)
       a) Maximum area in right of min value (not including the min)
       c) Maximum area including min */
    return max(getMaxAreaRec( n, l, m - 1),
               getMaxAreaRec( n, m + 1, r),
               (r - l + 1)*(hist[m]) );
  }
  
  // The main function to find max area
  static int getMaxArea( int n)
  {
    // Build segment tree from given array. This takes
    // O(n) time
    constructST(n);
  
    // Use recursive utility function to find the
    // maximum area
    return getMaxAreaRec( n, 0, n - 1);
  }
  
  // Driver program to test above functions
  public static void main(String[] args)
  {
    int[] a = {6, 1, 5, 4, 5, 2, 6};
    int n = a.length;
    hist = new int[n];
  
    hist = a;
    System.out.print("Maximum area is " +  getMaxArea(n));
  }
}
  
// This code is contributed by Rajput-Ji 

Python3

# Python3 program for range minimum  
# query using segment tree
  
# modified to return index of minimum instead of minimum itself
# for further reference link
# https://www.geeksforgeeks.org/segment-tree-set-1-range-minimum-query/
  
#-------------------------------------------------------------------------
from math import ceil,log2;  
    
# A utility function to get  
# minimum of two numbers  
def minVal(hist,x, y) :
    if x==-1:
        return y
    if y==-1:
        return x
    return x if (hist[x] < hist[y]) else y;  
    
# A utility function to get the  
# middle index from corner indexes.  
def getMid(s, e) : 
    return s + (e - s) // 2;  
    
""" A recursive function to get the  
minimum value in a given range  
of array indexes. The following  
are parameters for this function.  
    
    st --> Pointer to segment tree  
    index --> Index of current node in the  
        segment tree. Initially 0 is  
        passed as root is always at index 0  
    ss & se --> Starting and ending indexes  
                of the segment represented  
                by current node, i.e., st[index]  
    qs & qe --> Starting and ending indexes of query range """
def RMQUtil( hist,st, ss, se, qs, qe, index) : 
    
    # If segment of this node is a part  
    # of given range, then return  
    # the min of the segment  
    if (qs <= ss and qe >= se) : 
        return st[index];  
    
    # If segment of this node  
    # is outside the given range  
    if (se < qs or ss > qe) : 
        return -1;  
    
    # If a part of this segment  
    # overlaps with the given range  
    mid = getMid(ss, se);  
    return minVal(hist,RMQUtil(hist,st, ss, mid, qs,  
                          qe, 2 * index + 1),  
                  RMQUtil(hist,st, mid + 1, se, 
                          qs, qe, 2 * index + 2));  
    
# Return minimum of elements in range  
# from index qs (query start) to  
# qe (query end). It mainly uses RMQUtil()  
def RMQ( hist,st, n, qs, qe) :  
    
    # Check for erroneous input values  
    if (qs < 0 or qe > n - 1 or qs > qe) : 
        
        print("Invalid Input");  
        return -1;  
        
    return RMQUtil(hist,st, 0, n - 1, qs, qe, 0);  
    
# A recursive function that constructs  
# Segment Tree for array[ss..se].  
# si is index of current node in segment tree st  
def constructSTUtil(hist, ss, se, st, si) : 
    
    # If there is one element in array,  
    # store it in current node of  
    # segment tree and return  
    if (ss == se) : 
    
        st[si] = ss;  
        return st[si];  
    
    # If there are more than one elements,  
    # then recur for left and right subtrees  
    # and store the minimum of two values in this node  
    mid = getMid(ss, se);  
    st[si] = minVal(hist,constructSTUtil(hist, ss, mid, 
                                    st, si * 2 + 1), 
                    constructSTUtil(hist, mid + 1, se, 
                                    st, si * 2 + 2));  
        
    return st[si];  
    
"""Function to construct segment tree  
from given array. This function allocates  
memory for segment tree and calls constructSTUtil() 
to fill the allocated memory """
def constructST( hist, n) : 
    
    # Allocate memory for segment tree  
    
    # Height of segment tree  
    x = (int)(ceil(log2(n)));  
    
    # Maximum size of segment tree  
    max_size = 2 * (int)(2**x) - 1;  
     
    st = [0] * (max_size);  
    
    # Fill the allocated memory st  
    constructSTUtil(hist, 0, n - 1, st, 0);  
    
    # Return the constructed segment tree  
    return st;  
    
  
#----------------------------------------------------------------
  
# main program
# Python3 program using Divide and Conquer
# to find maximum rectangular area under a histogram
  
  
def max_area_histogram(hist):
    area=0
    #initialize area
      
    st = constructST(hist, len(hist))
    # construct the segment tree
      
    try:
        # try except block is generally used in this way
        # to suppress all type of exceptions raised.
          
        def fun(left,right):
              
        # this function "fun" calculates area
        # recursively between indices left and right
              
            nonlocal area
              
            # global area won't work here as
            # variable area is defined inside function
            # not in main().
               
            if left==right:
                return
            # the recursion has reached end
               
              
            index = RMQ(hist,st, len(hist), left, right-1)
            # RMQ function returns index 
            # of minimum value
            # in the range of [left,right-1]
            # can also be found by using min() but
            # results in O(n) instead of O(log n) for traversing
              
            area=max(area,hist[index]*(right-left))
            # calculate area with minimum above
              
            fun(index+1,right)
            fun(left,index)
            # initiate further recursion
               
            return
                   
        fun(0,len(hist))
        # initializes the recursion
          
        return(area)
        # return the max area to calling function
        # in this case "print"
           
    except:
        pass
      
# Driver Code 
hist = [6, 2, 5, 4, 5, 1, 6] 
print("Maximum area is",  
       max_area_histogram(hist)) 
    
# This code is contributed  
# by Vishnudev C.

C#

// C# code to implement the approach
using System;
using System.Numerics;
using System.Collections.Generic;
  
public class GFG {
  
  static int[] hist;
  static int[] st;
    
  // A utility function to find minimum of three integers
  static int max(int x, int y, int z)
  {  return Math.Max(Math.Max(x, y), z); }
  
  // A utility function to get minimum of two numbers in hist[]
  static int minVal(int i, int j)
  {
    if (i == -1) return j;
    if (j == -1) return i;
    return (hist[i] < hist[j])? i : j;
  }
  
  // A utility function to get the middle index from corner indexes.
  static int getMid(int s, int e)
  {   return s + (e -s)/2; }
  
  /*  A recursive function to get the index of minimum value in a given range of
    indexes. The following are parameters for this function.
  
    hist   -. Input array for which segment tree is built
    st    -. Pointer to segment tree
    index -. Index of current node in the segment tree. Initially 0 is
             passed as root is always at index 0
    ss & se  -. Starting and ending indexes of the segment represented by
                 current node, i.e., st[index]
    qs & qe  -. Starting and ending indexes of query range */
  static int RMQUtil( int ss, int se, int qs, int qe, int index)
  {
    // If segment of this node is a part of given range, then return the
    // min of the segment
    if (qs <= ss && qe >= se)
      return st[index];
  
    // If segment of this node is outside the given range
    if (se < qs || ss > qe)
      return -1;
  
    // If a part of this segment overlaps with the given range
    int mid = getMid(ss, se);
    return minVal( RMQUtil(ss, mid, qs, qe, 2*index+1),
                  RMQUtil( mid+1, se, qs, qe, 2*index+2));
  }
  
  // Return index of minimum element in range from index qs (query start) to
  // qe (query end).  It mainly uses RMQUtil()
  static int RMQ(  int n, int qs, int qe)
  {
    // Check for erroneous input values
    if (qs < 0 || qe > n-1 || qs > qe)
    {
      Console.Write("Invalid Input");
      return -1;
    }
  
    return RMQUtil(  0, n-1, qs, qe, 0);
  }
  
  // A recursive function that constructs Segment Tree for hist[ss..se].
  // si is index of current node in segment tree st
  static int constructSTUtil(int ss, int se, int si)
  {
    // If there is one element in array, store it in current node of
    // segment tree and return
    if (ss == se)
      return (st[si] = ss);
  
    // If there are more than one elements, then recur for left and
    // right subtrees and store the minimum of two values in this node
    int mid = getMid(ss, se);
    st[si] =  minVal( constructSTUtil( ss, mid,  si*2+1),
                     constructSTUtil( mid+1, se,  si*2+2));
    return st[si];
  }
  
  /* Function to construct segment tree from given array. This function
   allocates memory for segment tree and calls constructSTUtil() to
   fill the allocated memory */
  static void constructST(int n)
  {
    // Allocate memory for segment tree
    int x = (int)(Math.Ceiling(Math.Log(n))); //Height of segment tree
    int max_size = 2*(int)Math.Pow(2, x) - 1; //Maximum size of segment tree
    st = new int[max_size*2];
  
    // Fill the allocated memory st
    constructSTUtil( 0, n-1, 0);
  
    // Return the constructed segment tree
    // return st;
  }
  
  // A recursive function to find the maximum rectangular area.
  // It uses segment tree 'st' to find the minimum value in hist[l..r]
  static int getMaxAreaRec( int n, int l, int r)
  {
    // Base cases
    if (l > r)  return Int32.MinValue;
    if (l == r)  return hist[l];
  
    // Find index of the minimum value in given range
    // This takes O(Logn)time
    int m = RMQ( n, l, r);
  
    /* Return maximum of following three possible cases
       a) Maximum area in Left of min value (not including the min)
       a) Maximum area in right of min value (not including the min)
       c) Maximum area including min */
    return max(getMaxAreaRec( n, l, m - 1),
               getMaxAreaRec( n, m + 1, r),
               (r - l + 1)*(hist[m]) );
  }
  
  // The main function to find max area
  static int getMaxArea( int n)
  {
    // Build segment tree from given array. This takes
    // O(n) time
    constructST(n);
  
    // Use recursive utility function to find the
    // maximum area
    return getMaxAreaRec( n, 0, n - 1);
  }
  
// Driver Code
public static void Main(string[] args)
{
    int[] a = {6, 1, 5, 4, 5, 2, 6};
    int n = a.Length;
    hist = new int[n];
  
    hist = a;
    Console.WriteLine("Maximum area is " +  getMaxArea(n));
}
}
Producción

Maximum area is 12

Este problema se puede resolver en tiempo lineal. Vea a continuación el conjunto 2 para la solución de tiempo lineal. 
Solución de tiempo lineal para el área rectangular más grande en un histograma

Publicación traducida automáticamente

Artículo escrito por GeeksforGeeks-1 y traducido por Barcelona Geeks. The original can be accessed here. Licence: CCBY-SA

Deja una respuesta

Tu dirección de correo electrónico no será publicada. Los campos obligatorios están marcados con *