After All these Years!

[Namir]

Hi All HP-65 enthusiasts,

I think I have found a bug in the HP-65 Math Pac 2's program that calculates the Incomplete Gamma function. The listing on page 97 of the STAT PAC 2 needs the following three additional instructions, before the RTN instruction, to yield the correct answers:

CHS

1

+

I checked my corrected code on an HP41CX and the Alpha Wolfram web site. The corrections give results that match that on the Alpha Wolfram site.

I also noticed something funny in the listing of the Gamma function on page 96. The HP-65 program uses two cards. The program in the first card initializes the constants used by the second card's program which performs the gamma function calculations. However, the listing of Card 2 re-keys one of the constants instead of recalling its value that should already be in a memory register. Looks like sloppy programming to me!!!

Has anyone else found other weird listing issues with the HP-65 STAT PAC 2? I know that this is very old software and the issue should be addressed with the British Museum or the Louvre Museum instead of HP.

:-)

Namir

Edited: 24 May 2012, 10:06 a.m.

[Matt Agajanian]

Namir (or should I call you Hercule Poirot), well done! What baffles me is why, during the 65's lifetime (or during the original users' lifetimes), this glitch went unnoticed. You'd think, with the 65 being the first handheld computer as well as the plethora of scientists, mathematicians, engineers, and even statisticians who originally bought the 65 in its first run would've spotted the algorithm's errors firsthand.

Once again I say, well done!

[Walter B]

Another explanation may be nobody used that Incomplete Gamma program ;-)

And that 'sloppy programming' for Gamma didn't cause any errors. So what?

[Namir]

I agree with you on both issues.

Regarding the Gamma function, the correct program would be several steps shorter if one replaced the steps used to enter the constant with a RCL command. Given that the HP-65 was not well endowed with memory like the HP-67 and the HP-41C, the shorter version should be more welcome.

Namir

[Namir]

Thank you Matt! I recently bought the HP-65 Math Pac 2 and was curious about the various programs that calculated advanced functions. I was curious, by sheer chance, about the implementation of incomplete gamma function. The short code in the manual raised my curiosity, so I entered the code in an HP41CX and noticed that the results did not match the test cases results (which gave the correct answers) in the manua.

Edited: 24 May 2012, 3:48 p.m.

[Matt Agajanian]

You're welcome.