Approximate zero values in an array

[David Skelly]

It's not really the average, though, if I've understood the question. The sequence 23, 0, 20 becomes 23, 21.5, 20. 18, 0, 0, 24 becomes 18, 20, 22, 24 20, 0, 0, 0, 16 becomes 20, 19, 18, 17, 16. Not sure what the technical term for that is. Interpolation? Even so, it is still simple maths. The only tricky bit is handling the end conditions if a 0 occurs at the first or last position in the array.

[Lost User]

David Skelly wrote:

It's not really the average, though, if I've understood the question.

You're right of course, I was confused (quite easy thing to do) by the questioner using the expression "approximate(average)".

[Lost User]

David Skelly wrote:

Interpolation

Ding ding ding! Winner winner chicken dinner!

Check out the CodeProject forum Guidelines[^] The original soapbox 1.0 is back![^]

[Johnkokk]

Ok, we established that what i need, is some sort of linear interpolation. Any ideas ? :)

[nlarson11]

    Dim arr() As Single = {17, 18, 0, 0, 24, 23, 0, 20, 0, 0, 0, 16}
    Dim idx1 As Int16 = -1, idx2 As Int16 = -1
    Dim FoundAZero As Boolean = False

    For i As Int16 = 0 To arr.GetUpperBound(0)

        If Not FoundAZero Then
            If arr(i) = 0 Then
                FoundAZero = True 'once we find an zero the else will be used below until we find the next non zero 
            Else
                idx1 = i 'hold on to the index every non zero
            End If
        Else
            If arr(i) <> 0 Then '
                idx2 = i 'hold on to the index of the non zero 
                '--------------------------------------
                'get the value that's held in both indexes
                Dim val1 As Int16 = arr(idx1), val2 As Int16 = arr(idx2)
                Dim IncVal As Single

                'increment value is non zero 1 minus non zero 2 / the number of zeros between
                'the \* -1 is whether or not we need to increment up or down depending on values of the array

                If val1 < val2 Then
                    IncVal = (val2 - val1) / (idx2 - idx1)
                Else
                    IncVal = ((val1 - val2) / (idx2 - idx1)) \* -1
                End If

                For x As Int16 = idx1 + 1 To idx2 - 1
                    arr(x) = arr(x - 1) + IncVal 'replace the 0 inbetween both indexes with the incremented value 
                Next
                '--------------------------------------
                idx1 = idx2        'take the last none zero and set it to the first hold as we continue down the array
                FoundAZero = False 'reset finding a zero flag
            End If
        End If
    Next

'Never argue with an idiot; they'll drag you down to their level and beat you with experience.' ~ anonymous 'Life's real failure is when you do not realize how close you were to success when you gave up.' ~ anonymous

[Johnkokk]

Ok, the code above works almost perfectly. THe only problem is that if the first position of the array is populated with 0, it crashes at " Dim val1 As Int16 = arr(idx1)", because idx1 remains -1 Thanks.

[nlarson11]

I understood that when I wrote but I figured it wouldn't be possible to have the array start with a zero or end with one because how can you figure out what those should be. I guess you could take the number to the right of the 0 and divide by 2 but wasn't sure if that's what should happen.

'Never argue with an idiot; they'll drag you down to their level and beat you with experience.' ~ anonymous 'Life's real failure is when you do not realize how close you were to success when you gave up.' ~ anonymous

[peterchen]

Hmmmmm... chicken!

Personally, I love the idea that Raymond spends his nights posting bad regexs to mailing lists under the pseudonym of Jane Smith. He'd be like a super hero, only more nerdy and less useful. [Trevel]
| FoldWithUs! | sighist | µLaunch - program launcher for server core and hyper-v server

[Gideon Engelberth]

You may want to turn on Option Strict. When you use val1 and val2 as integers, you will produce incorrect results if one of the non-zero values has a decimal component. Also, I would recommend using Integer instead of Int16 for the index variables as turning on Option Strict will cause errors for both of your loop declarations.

[Gideon Engelberth]

Then add some error checking. Assuming you put this code in a method, you can do the check at the right at the top of the method to make sure that the parameter a) is not nothing, b) does not start with a zero, and c) does not end with a zero and throw the appropriate exception. Like so:

Private Sub Interpolate(ByVal values() As Single)
If values Is Nothing Then Throw New ArgumentNullException("values")
'nothing to do in this routine and the other error checks would cause errors
If values.Length = 0 Then Exit Sub
If values(0) = 0 Then
Throw New ArgumentException("Value array cannot start with a 0.", "values")
End If
If values(values.Length - 1) = 0 Then
Throw New ArgumentException("Value array cannot end with a 0.", "values")
End If

'other code to do interpolation here

End Sub