# still probs with JD algorithm

4 posts / 0 new
pox
still probs with JD algorithm

Hi,

Thought I'd cracked it, but the algorithm below (VB) today produced a JD result that was one day less than the true version. Nothing to do with pre-noon because I tried this at 5.00 pm local time (which is Greenwich anyway!)

Public Function julian(Y As Integer, m As Integer, d As Integer) As Long
Dim a As Integer, b As Integer
a = Int(Y / 100)
b = 2 - a + Fix(a / 4)
If m > 2 Then
'do nothing
Else
Y = Y - 1
m = m + 12
End If
julian = Int(365.25 * (Y + 4716)) + Fix(30.6001 * (m + 1)) + d + b - 1524.5
End Function

(I have used 'fix' where the value inside the function could be negative even though in a practical sense today it couldn't be, so fix(-2.4) gives -2 as output.  What could I be doing wrong?

OK - I don't think the problem is with the algorithm, it's what I was doing after the function ran. I was subtracting 0.5 day from the result whereas if I add 0.5d the result comes out correctly. A note for the unwary maybe?

David Benn
Meeus JD implementation

Hi Michael

If it helps, here's an implementation of calendar date to JD from Meeus, 1991, Astronomical Algorithms, ch 7:

https://sourceforge.net/p/vstar/code/1453/tree/tags/DEV-AAVSO-06May2017/src/org/aavso/tools/vstar/util/date/MeeusDateUtil.java

starting around line 97, for comparison with what you're doing (preserving Meeus's terse variable names in most cases), along with a bunch of unit tests:

https://sourceforge.net/p/vstar/code/1453/tree/tags/DEV-AAVSO-06May2017/test/org/aavso/tools/vstar/util/date/DateUtilTestBase.java

The directories in which those source files exist also contain HJD conversion source (code, test).

Addition: it looks like you may be using Meeus as a reference as well, directly or indirectly. The main differences appear to be that I'm checking for whether a date is before or after the Gregorian Calendar came into effect and the slightly different rounding in the last line of the function.

David

pox
Hi David,

Hi David,

Yes, and of course for my purposes it's purely non-past dates that matter, therefore the 'easy' Meeus. The problem I had with some of his online refs was the text quality of the scanned article - it was hard to tell +, * and - apart sometimes! Actually used the updated/corrected version with my software tonight, and everything now fine. Thanks for your help.

David Benn
No worries

No worries Michael.

David