Improve partitioning algorithm

Question

I was trying to write an algorithm that builds an array of dates ranges using a start and end date and a delta.

In broader terms, the algorithm input and output are as following :

/*
    INPUT : T1 of type DateTimeoffset.
            T2 of type DateTimeoffset.
            delta of type Int.
        
    OUTPUT : An array of time ranges [x, y] where x is start date and y is end date.
    
    R1 = [T1             , T1 + delta * 1]
    R2 = [T1 + delta * 1 , T1 + delta * 2]
    R3 = [T1 + delta * 2 , T1 + delta * 3]
    R4 = [T1 + delta * 3 , T1 + delta * 4]
    R5 = [T1 + delta * 4 , T1 + delta * 5]
    R6 = [T1 + delta * 5 , T2]
*/  

The C# code I built for this is as following:

public static (DateTimeOffset From, DateTimeOffset To)[] Partition(DateTimeOffset start, DateTimeOffset end, int delta)
{
    var ranges = new List<(DateTimeOffset From, DateTimeOffset To)>();
    var lower = start;

    var offset = 0;

    while (lower < end)
    {
        var lowerbound = start + TimeSpan.FromMinutes(delta * offset);
        var upperbound = start + TimeSpan.FromMinutes(delta * ++offset);

        if(upperbound > end)
            upperbound = end;

        ranges.Add((From: lowerbound, To: upperbound));

        lower = upperbound;
    }

    return ranges.ToArray();
}

is there a better way to write this algorithm and if not how can it be improved ?

Answer 1

Seems like you don't need to multiply the offset each time, just keep track of the previous upper bound:

public static (DateTimeOffset From, DateTimeOffset To)[] Partition(DateTimeOffset start, DateTimeOffset end, int delta)
{
    var ranges = new List<(DateTimeOffset From, DateTimeOffset To)>();
    var lowerbound = start;
    
    while (lowerbound < end)
    {
        var upperbound = lower + TimeSpan.FromMinutes(delta);
        
        if(upperbound > end) upperbound = end;
        
        ranges.Add((From: lowerbound, To: upperbound));
        
        lowerbound = upperbound;
    }
    
    return ranges.ToArray();
}

Other options:

• Do you need to return an array? Could you just return the list or an IEnumerable and avoid the (relatively cheap) conversion to an array?
• You could use yield to return values as they are computed, but the only benefit is if a consumer doesn't need the entire list and can stop the generation before it reaches the end.