共用方式為


BigInteger.Min(BigInteger, BigInteger) 方法

定義

傳回兩個 BigInteger 值的較小值。

public:
 static System::Numerics::BigInteger Min(System::Numerics::BigInteger left, System::Numerics::BigInteger right);
public:
 static System::Numerics::BigInteger Min(System::Numerics::BigInteger left, System::Numerics::BigInteger right) = System::Numerics::INumber<System::Numerics::BigInteger>::Min;
public static System.Numerics.BigInteger Min (System.Numerics.BigInteger left, System.Numerics.BigInteger right);
static member Min : System.Numerics.BigInteger * System.Numerics.BigInteger -> System.Numerics.BigInteger
Public Shared Function Min (left As BigInteger, right As BigInteger) As BigInteger

參數

left
BigInteger

要比較的第一個值。

right
BigInteger

要比較的第二個值。

傳回

leftright 參數,擇一較小者。

實作

範例

下列範例會 Min 使用 方法來選取值數位 BigInteger 中的最小數位。

using System;
using System.Numerics;

public class Example
{
   public static void Main()
   {
      BigInteger[] numbers = { Int64.MaxValue * BigInteger.MinusOne,
                               BigInteger.MinusOne,
                               10359321239000,
                               BigInteger.Pow(103988, 2),
                               BigInteger.Multiply(Int32.MaxValue, Int16.MaxValue),
                               BigInteger.Add(BigInteger.Pow(Int64.MaxValue, 2),
                                              BigInteger.Pow(Int32.MaxValue, 2)),
                               BigInteger.Zero };
      if (numbers.Length < 2)
      {
         Console.WriteLine("Cannot determine which is the smaller of {0} numbers.",
                            numbers.Length);
         return;
      }

      BigInteger smallest = numbers[numbers.GetLowerBound(0)];

      for (int ctr = numbers.GetLowerBound(0) + 1; ctr <= numbers.GetUpperBound(0); ctr++)
         smallest = BigInteger.Min(smallest, numbers[ctr]);

      Console.WriteLine("The values:");
      foreach (BigInteger number in numbers)
         Console.WriteLine("{0,55:N0}", number);

      Console.WriteLine("\nThe smallest number of the series is:");
      Console.WriteLine("   {0:N0}", smallest);
   }
}
// The example displays the following output:
//       The values:
//                                    -9,223,372,036,854,775,807
//                                                            -1
//                                            10,359,321,239,000
//                                                10,813,504,144
//                                            70,366,596,661,249
//            85,070,591,730,234,615,852,008,593,798,364,921,858
//                                                             0
//
//       The smallest number of the series is:
//          -9,223,372,036,854,775,807.
open System
open System.Numerics

let numbers =
    [| bigint Int64.MaxValue * BigInteger.MinusOne
       BigInteger.MinusOne
       10359321239000I
       BigInteger.Pow(103988I, 2)
       BigInteger.Multiply(Int32.MaxValue, Int16.MaxValue)
       BigInteger.Add(BigInteger.Pow(Int64.MaxValue, 2), BigInteger.Pow(Int32.MaxValue, 2))
       BigInteger.Zero |]

if numbers.Length < 2 then
    printfn $"Cannot determine which is the smaller of {numbers.Length} numbers."
else
    let mutable smallest = numbers[0]

    for ctr = 1 to numbers.Length - 1 do
        smallest <- BigInteger.Min(smallest, numbers[ctr])

    printfn "The values:"

    for number in numbers do
        printfn $"{number, 55:N0}"

    printfn "\nThe smallest number of the series is:"
    printfn $"   {smallest:N0}"
// The example displays the following output:
//       The values:
//                                    -9,223,372,036,854,775,807
//                                                            -1
//                                            10,359,321,239,000
//                                                10,813,504,144
//                                            70,366,596,661,249
//            85,070,591,730,234,615,852,008,593,798,364,921,858
//                                                             0
//
//       The smallest number of the series is:
//          -9,223,372,036,854,775,807.
Imports System.Numerics

Module Example
   Public Sub Main()
      Dim numbers() As BigInteger = { Int64.MaxValue * BigInteger.MinusOne, 
                                      BigInteger.MinusOne, 
                                      10359321239000, 
                                      BigInteger.Pow(103988, 2),
                                      BigInteger.Multiply(Int32.MaxValue, Int16.MaxValue), 
                                      BigInteger.Add(BigInteger.Pow(Int64.MaxValue, 2), 
                                                     BigInteger.Pow(Int32.MaxValue, 2)),
                                      BigInteger.Zero }
      If numbers.Length < 2 Then 
         Console.WriteLine("Cannot determine which is the smaller of {0} numbers.",
                            numbers.Length)
         Exit Sub
      End If
           
      Dim smallest As BigInteger = numbers(numbers.GetLowerBound(0))
      
      For ctr As Integer = numbers.GetLowerBound(0) + 1 To numbers.GetUpperBound(0)
         smallest = BigInteger.Min(smallest, numbers(ctr))
      Next
      Console.WriteLine("The values:")
      For Each number As BigInteger In numbers
         Console.WriteLine("{0,55:N0}", number)
      Next   
      Console.WriteLine()
      Console.WriteLine("The smallest number of the series is:")
      Console.WriteLine("   {0:N0}", smallest)   
   End Sub
End Module
' The example displays the following output:
'       The values:
'                                    -9,223,372,036,854,775,807
'                                                            -1
'                                            10,359,321,239,000
'                                                10,813,504,144
'                                            70,366,596,661,249
'            85,070,591,730,234,615,852,008,593,798,364,921,858
'                                                             0
'       
'       The smallest number of the series is:
'          -9,223,372,036,854,775,807.

備註

這個方法會對應至 Math.Min 基本數值類型的方法。

適用於

另請參閱