HP-25 statistics weirdness

[John Garza (3665)]

Ever notice the HP-25 does not accumulate the y^2 value in a stat register? Makes things like computing r^2 for linear regression a chore as you have to account for it in your code on data entry. Also, RCL Sigma+ is not allowed - making our fearless pilot Mr. Sweeny hit a couple of extra keys - not to mention having to remember where SigmaX and SigmaY are stored.

Doing a little research, I found the HP-25 is unique in this. It's predecessor, the HP-55 has both y^2 and RCL Sigma+, and so does it's successors, the HP29C and the HP-33E/C. The HP-27 has y^2 but lacks RCL Sigma+.

I presume all this weirdness is due to space imitations? And yet the 25 has percentage and NOP ?

-J

[Thomas Klemm]

I could imagine that \(\sum y\) and \(\sum xy\) can still be useful in case of a frequency table.
You can interpret \(y\) as frequency and \(x\) as value.

The mean is then:

\(
\bar{x} = \frac{\sum xy}{\sum y}
\)

Example:
\(
\begin{array}{|c|c|}
\hline
Frequency & Value \\
\hline
3 & 7 \\
5 & 11 \\
2 & 13 \\
\hline
\end{array}
\)

CLEAR REG
3 ENTER 7 \(\Sigma +\)
5 ENTER 11 \(\Sigma +\)
2 ENTER 13 \(\Sigma +\)

RCL 5
RCL 4
\(\div\)

10.20

[John Garza (3665)]

BTW,

How did you get the sigma symbols to display in your post?
I looked and could not find that in the editor.

-J

[Thomas Klemm]

You can hit the button on any post to see the code.
In this case it is embedded LaTeX, or specifically \sum for the \(\sum\) sign.
Or then \Sigma + for \(\Sigma +\).

[Steve Simpkin]

(10-09-2023 12:27 PM)John Garza (3665) Wrote:  Ever notice the HP-25 does not accumulate the y^2 value in a stat register? Makes things like computing r^2 for linear regression a chore as you have to account for it in your code on data entry. Also, RCL Sigma+ is not allowed - making our fearless pilot Mr. Sweeny hit a couple of extra keys - not to mention having to remember where SigmaX and SigmaY are stored.

Doing a little research, I found the HP-25 is unique in this. It's predecessor, the HP-55 has both y^2 and RCL Sigma+, and so does it's successors, the HP29C and the HP-33E/C. The HP-27 has y^2 but lacks RCL Sigma+.

I presume all this weirdness is due to space imitations? And yet the 25 has percentage and NOP ?

-J

I believe you are correct about why that feature was not included. As William C Wickes has said on other HP calculator projects when asked why a feature was missing from a particular model, "Life is short and ROM is full". In the case of the HP-25, this is literally true. There was no room left in the HP-25's ROM to add *any* feature unless you removed a feature or were able to further optimize the code to make more room.
https://www.hpmuseum.org/forum/thread-85...l#pid74865

I suspect they put a higher priority on the percentage and NOP functions than on accumulating the y^2 value in a statistics register. Considering the HP-25 didn't start out as a programmable calculator, I think they did an amazing job including the features they did in the very limited ROM space they had available.