Excel's MOD in C#?

Peter_in_2780

Modulo and negative numbers make an explosive mix. See the last table in Modulo operation - Wikipedia[^]. One common trick is to do something like

result = (a + n*b) % b

where n is some "sufficiently large" integer. Cheers, Peter

Software rusts. Simon Stephenson, ca 1994. So does this signature. me, 2012

OriginalGriff

MOD in Excel is a true Modulus: % in C# is a remainder, and the two are not the same: Mod and Remainder are not the Same – Rob Conery[^] What you need is to write a modulus method:

    private double Mod(double a, double b)
        {
        return a - b \* Math.Floor(a / b);
        }

Call that, and you'll get the answer you're expecting.

Sent from my Amstrad PC 1640 Never throw anything away, Griff Bad command or file name. Bad, bad command! Sit! Stay! Staaaay... AntiTwitter: @DalekDave is now a follower!

Jassim Rahma

double bearing4 = (-1.779406909 + 2 * Math.PI) % Math.PI;

It returns 1.362 which is different that Excel

www.softnames.com

Jassim Rahma

Yes Yes Yes Thank you so much

www.softnames.com

Peter_in_2780

Jassim Rahma wrote:

which is different that Excel

which is because you're calculating something different.

Software rusts. Simon Stephenson, ca 1994. So does this signature. me, 2012

OriginalGriff

You're welcome!

Sent from my Amstrad PC 1640 Never throw anything away, Griff Bad command or file name. Bad, bad command! Sit! Stay! Staaaay... AntiTwitter: @DalekDave is now a follower!

BillWoodruff

Also useful background: Eric Lippert "What’s the difference? Remainder vs Modulus" [^]

«Where is the Life we have lost in living? Where is the wisdom we have lost in knowledge? Where is the knowledge we have lost in information?» T. S. Elliot

BillWoodruff

what I use:

using System;
using System.Runtime.InteropServices.WindowsRuntime;

namespace Utilities
{
public static class MathExtensions
{
// ideas from: Lippert, Gravell, Skeet
public static double ModT(this T i1, T i2)
{
double v1 = i1.GetDouble();
double v2 = i2.GetDouble();
return v1 - v2 * Math.Floor(v1 / v2);
}

    public static double ModT1T2(this T1 i1, T2 i2)
    {
        double v1 = i1.GetDouble();
        double v2 = i2.GetDouble();
        return v1 - v2 \* Math.Floor(v1 / v2);
    }

    public static double GetDouble(this T i1)
    {
        double v1;

        try
        {
            v1 = Convert.ToDouble(i1);
        }
        catch (InvalidCastException iex)
        {
            throw new InvalidCastException($"type {typeof(T)} cannot be cast to double: {iex.Message}");
        }

        catch (FormatException fex)
        {
            throw new FormatException($"type {typeof(T)} cannot be used: {fex.Message}");
        }

        catch (OverflowException oex)
        {
            throw new OverflowException($"type {typeof(T)} result is an overflow: {oex.Message}");
        }

        return v1;
    }
}

}

Tests:

        double dbl = -1.779406909.ModT(2 \* Math.PI);

        double int1 = 221.ModT(20);

        double sngl = 102.0f.ModT(20.0f);

        double deci = 102.0m.ModT(20.0m);

        double intdecicombo = 100.0.ModT1T2(34.56m);

        double dbl2 = 100.0.ModT(34.56);

«Where is the Life we have lost in living? Where is the wisdom we have lost in knowledge? Where is the knowledge we have lost in information?» T. S. Elliot

Richard Deeming

BillWoodruff wrote:

catch (InvalidCastException iex)
{
throw new InvalidCastException($"type {typeof(T)} cannot be cast to double: {iex.Message}");
}

If you're going to wrap an exception, you should pass the wrapped exception as the inner exception. You should probably review the error messages as well. For example, it might be fine to use the given type, but not the specified value.

catch (InvalidCastException iex)
{
throw new InvalidCastException($"Type {typeof(T)} cannot be cast to double: {iex.Message}", iex);
}
catch (FormatException fex)
{
throw new FormatException($"Value '{i1}' of type {typeof(T)} cannot be converted to a double: {fex.Message}", fex);
}
catch (OverflowException oex)
{
throw new OverflowException($"Value '{i1}' of type {typeof(T)} is too large for a double: {oex.Message}", oex);
}

You can also eliminate the InvalidCastException by adding a generic type constraint, since the documentation[^] says it's only thrown if the value to convert doesn't implement IConvertible:

public static double ModT<T>(this T i1, T i2) where T : IConvertible { ... }
public static double ModT1T2<T1, T2>(this T1 i1, T2 i2) where T1 : IConvertible where T2 : IConvertible { ... }
public static double GetDouble<T>(this T i1) where T : IConvertible { ... }

"These people looked deep within my soul and assigned me a number based on the order in which I joined." - Homer

BillWoodruff

Well, there is no code I write you cannot improve, and I am happy to see the improvements, no matter how ephemeral :) You are already aware, I'm sure, of the long-standing issue in constraining generic parameters to numeric Types. If you're going to constrain, might as well go for broke:

 where T:
      struct, 
      IComparable, 
      IComparable, 
      IConvertible, 
      IEquatable, 
      IFormattable

Note: var c1dbl = 'z'.ModT('a'); // invalid cast error Of course, that's somewhat ironic :) since you can do this: int c1int = 'z' % 'a'; // returns #25

«Where is the Life we have lost in living? Where is the wisdom we have lost in knowledge? Where is the knowledge we have lost in information?» T. S. Elliot

Richard Deeming

That's an annoying restriction on the char type. All of the floating-point IConvertible members throw an InvalidCastException. But casting to a floating-point type, either explicitly or implicitly, works. :doh:

char c = '*';
double a = (double)c; // Works
double b = Convert.ToDouble(c); // Throws

Reference Source[^]

"These people looked deep within my soul and assigned me a number based on the order in which I joined." - Homer

BillWoodruff

Would it be the language we love without these quirks :omg: Seems to me we could have had INumericType ... IIntType ... IFloatType constraints ... by now ... given the frequency of posts about the necessity for these. cheers, Bill

«Where is the Life we have lost in living? Where is the wisdom we have lost in knowledge? Where is the knowledge we have lost in information?» T. S. Elliot

Richard Deeming

Looks like it's part of a much larger discussion that's going to take a while to make it into the language (if it ever does): Champion "Type Classes (aka Concepts, Structural Generic Constraints)" · Issue #110 · dotnet/csharplang · GitHub[^] Meanwhile, you can "fake" it by using Jon Skeet's generic operators from the MiscUtil project: Generic Operators[^] It hasn't been updated since 2009, but the concept still works. For example:

using System.Linq.Expressions;

public static class GenericOperators<T>
{
public static readonly Func<T, T, T> Add = Create(Expression.Add);
public static readonly Func<T, T, T> Subtract = Create(Expression.Subtract);
public static readonly Func<T, T, T> Multiply = Create(Expression.Multiply);
public static readonly Func<T, T, T> Divide = Create(Expression.Divide);

private static Func<T, T, T> Create(Func<Expression, Expression, BinaryExpression> body)
{
    try
    {
        Type typeT = typeof(T);
        var left = Expression.Parameter(typeT, "left");
        var right = Expression.Parameter(typeT, "right");

        if (typeT.IsEnum)
        {
            Type enumType = Enum.GetUnderlyingType(typeT);
            var x = Expression.Convert(left, enumType);
            var y = Expression.Convert(right, enumType);

            Expression op = body(x, y);
            if (op.Type == enumType) op = Expression.Convert(op, typeT);

            return Expression.Lambda<Func<T, T, T>>(op, left, right).Compile();
        }

        return Expression.Lambda<Func<T, T, T>>(body(left, right), left, right).Compile();
    }
    catch (InvalidOperationException ex)
    {
        string message = ex.Message;
        return delegate { throw new InvalidOperationException(message); };
    }
    catch (ArgumentException ex)
    {
        string message = ex.Message;
        return delegate { throw new InvalidOperationException(message); };
    }
}

}

Of course, it wouldn't help much in this case, because you'd still n