Calculator benchmark - Printable Version +- HP Forums (https://www.hpmuseum.org/forum) +-- Forum: Not HP Calculators (/forum-7.html) +--- Forum: Not remotely HP Calculators (/forum-9.html) +--- Thread: Calculator benchmark (/thread-17338.html) |
RE: Calculator benchmark - robve - 08-12-2021 11:37 PM SHARP PC-G850(V)(S): 107 keystrokes Same number of keystrokes as the PC-E500(S) but one more for R. instead of F1 to run. The SHARP PC-G850(V)(S) has the same or similar keys (but no function keys) and similar BASIC. SHARP PC-1403 to PC-1475: 118 keystrokes An older series with the same or similar keys as the PC-E500(S), but with simpler BASIC, which has no impact on keying-in this simple program. However, [ENTER] is required after each pair of X,Y results are displayed (because all PC-14xx have a one-line LCD, except PC-1475 which has two lines. Thus adding 11 more keystrokes compared to the PC-E500(S). Edit note: by comparison, the AREAD-based program for the PC-14xx series takes 111 keystrokes, here in full detail: Code: [MODE] 1 PRO mode Code: 3.12[DEF]A[CL]2.56[DEF]B 13 Total score: 111 Total score is 110 when entering B first before A for the first entry pair. This avoids having to clear the error [CL] when B is zero. - Rob RE: Calculator benchmark - robve - 08-13-2021 02:16 PM (08-08-2021 05:42 AM)jhallen Wrote: I've written a silly benchmark for my growing calculator collection- it measures the number of keystrokes necessary to transform a table of numbers through some equations. The results are here: Just my 2c: the keystrokes for the Casio fx-cg50 aren't the same as Casio fx-9750GIII (I own the former, not the latter). There is no Shift Setup F2 to switch to line I/O mode (it selects COORD instead) and EXE works differently. Use Shift Prgm and keep the FN menu to enter Prgm symbols. AEXE inserts a newline/carriage return [NL] symbol for the next statement: Code: Shift Prgm Activate Prgm FN keys Program: 27 keystrokes. Then EXIT EXIT EXE to run the program. Execution: all 24 values must be entered (99) each with EXE (24), then another 12 EXE to display the second value, and 12 EXE + 11 EXE to rerun = 158 keystrokes, plus the 3 to exit the editor and run the program. Because there is no loop in the program, execution requires one more EXE after the second display to rerun the program. Adding a loop: Code: Shift Prgm Activate Prgm FN keys Program: 36 keystrokes. Execution: 147 keystrokes (99+24+24) plus the 3 (EXIT EXIT EXE) to exit the editor and run the program. Casio fx-CG50: 186 keystrokes. - Rob RE: Calculator benchmark - robve - 08-13-2021 09:10 PM (08-08-2021 05:42 AM)jhallen Wrote: Calculator Benchmark Another improvement. The Ti-89 Titanium and Voyage 200 are very similar. So I tried the Ti-89 program on the Voyage 200 starting from the usual Calc/Home screen. Program: Code: Apps Left Down Enter 3 New program The Voyage 200 has a Qwerty keyboard and a [ln] key that enters ln(. The setup takes 44 keystrokes. Execution: Code: 3.12 Enter Note that you don't have to always enter the values. Also 2x, y/2, and 2y work. Execution takes 77 keystrokes Voyage 200: 121 keystrokes It's a very nice CAS/graphing calculator and easy to program with the Qwerty keyboard. But graphing is a bit slow. The same x, y, 2x and y/2 input tricks can be used on the Ti-89, I assume. - Rob RE: Calculator benchmark - C.Ret - 08-15-2021 01:13 PM HP Prime [] and ( ) stand respectively for keystrokes and screentouches. SETUP: Code: [Apps](Spreadsheet) 2 Start a new Spreadsheet application Total 28 keystrokes and screentouches for setup. DATA entries and results display: Code: " 3.12 2.56" Already displayed 0 Total 65 keystrokes and screentouches for data entry and results display. Score 93 keystrokes and screen touches. RE: Calculator benchmark - robve - 08-15-2021 02:31 PM (08-15-2021 01:13 PM)C.Ret Wrote: Score 93 keystrokes and screentouches. Very nice!! The HP Prime is well designed and engineered. A touchscreen obviously offers a huge advantage with spreadsheets though. I had considered and briefly tried the spreadsheet app on Voyage 200, but quickly realized it didn't reduce the keystroke effort at all. Navigation takes longer with the cursor keys. Entering a value takes 7 or 8 keystrokes (Up/Down Enter 1.28 Enter). Editing a value in a cell takes more effort too, e.g. to change it to half its value takes 5 keystrokes (Enter Right / 2 Enter), same to double its value. And it's worse with the Casio fx-CG50 spreadsheet. Sadly, the Excel Android app isn't as good as the HP Prime Spreadsheet for this exercise. My Android device (Gemini) isn't a calculator, so it wouldn't count anyway. - Rob RE: Calculator benchmark - Valentin Albillo - 08-15-2021 03:47 PM . Hi, all What's the best solution (less keystrokes) so far for the HP-15C ? Please post exact code and keystroke count. And what's the least number of keystrokes among all models ? Please post model and keystroke count, Thanks in advance. V. RE: Calculator benchmark - robve - 08-15-2021 04:43 PM (08-15-2021 03:47 PM)Valentin Albillo Wrote: . I suspect that jhallen will update the table on GitHub eventually, that would be nice. If we want the absolute minimum keystrokes, then it may be premature to list the full table in this thread, because the totals are still changing by leaps (HP PRIME) or by small improvements. For example, on the PC-E500(S) we can use PF1 instead of R. for RUN. This saves one keystroke. I've updated my earlier post accordingly with an Edit 3. - Rob RE: Calculator benchmark - Maximilian Hohmann - 08-15-2021 05:16 PM Hello! (08-15-2021 04:43 PM)robve Wrote: I suspect that jhallen will update the table on GitHub eventually, that would be nice. Let me remind you of one crucial sentence in jhallen's GitHub site: "This calculator benchmark measures how fast it makes you, not how fast the calculator works internally" (my emphasis). So optimizing for the absolute minimum number of keystrokes is, in my understanding, counterproductive. Optimising can take an entire sunday afternoon whereas just typing ahead, as jhallen obviously did when he compiled his table, is done in a matter of seconds even if some keystrokes are wasted in the process. Regards Max RE: Calculator benchmark - robve - 08-15-2021 06:30 PM (08-15-2021 05:16 PM)Maximilian Hohmann Wrote: Let me remind you of one crucial sentence in jhallen's GitHub site: "This calculator benchmark measures how fast it makes you, not how fast the calculator works internally" (my emphasis). Right! I respectfully agree and would think this is true for pretty much all of the cases contributed by folks to this thread so far, no? The solution methods to key-in should be as straightforward as possible, perhaps earning points for simplicity rather than craftiness that requires a certain amount of mental agility that slows you down. However, besides keystrokes that take a fixed amount of time, what alternatives do you suggest to quantify "how fast it makes you"? - Rob RE: Calculator benchmark - Maximilian Hohmann - 08-15-2021 11:19 PM Hello! (08-15-2021 06:30 PM)robve Wrote: However, besides keystrokes that take a fixed amount of time, what alternatives do you suggest to quantify "how fast it makes you"? I would do it with a stopwatch. Count the seconds from turning on the calculator to getting the last result. Regards Max RE: Calculator benchmark - robve - 08-16-2021 01:01 AM (08-15-2021 11:19 PM)Maximilian Hohmann Wrote: I would do it with a stopwatch. Count the seconds from turning on the calculator to getting the last result. I'm afraid this is going nowhere. There are serious problems with this. When properly prepared and experienced with the calculator at hand, each keystroke should take a just about a fixed amount of time by a human to operate and register for this simple benchmark. If keystrokes are not quick or easy to make, because of the way the calculator works, then you are back to the problem you stated: "not how fast the calculator works internally", but that is exactly what you could end up timing. You can't have it both ways. The other problem is that time, as a quantitive measure of performance of a calculator operated by a human, is statistically and subjectively noisy. You will first need to establish a fair measure of timing to begin with. You can either take the fastest time among many users of the calculator, but then you're back at my first point, or you can determine the mean and deviation of the time to operate the calculator using a little army of human testers. The latter follows the traditional scientific method to ensure reproducibility of the experiment. It seems to me that all of this is totally out of scope for a simple fun benchmark experiment. Therefore, stick to the number of keystrokes as unbiased and reproducible. For what counts as a straightforward method, I will only speak for my examples and improvements, which are very basic and easy programs with executions that do not use any specialized or uncommon features of a calculator. For example, the PC-E500 program and execution are typical for using this BASIC calculator, requiring nothing special. Furthermore, unnecessary keystrokes should not count. For example, it is incorrect to rank Voyage 200 at 175 keystrokes when a normal straightforward method to solve this problem requires 121 keystrokes, simply by avoiding unnecessary keystrokes. - Rob RE: Calculator benchmark - EdS2 - 08-16-2021 06:22 AM . Good questions Valentin... (08-15-2021 03:47 PM)Valentin Albillo Wrote: What's the best solution (less keystrokes) so far for the HP-15C ? Please post exact code and keystroke count. That seems to be 104, as seen in post #3 upthread: (08-08-2021 07:26 PM)C.Ret Wrote: I found a way to spare keystrokes on HP-15C: Quote:And what's the least number of keystrokes among all models ? Please post model and keystroke count,That I don't know. As noted, one can hope that the results table here will be updated as progress is made. (Presently it shows the 48G at the top, with a score of 107, and if the result above is right, that can be beat by the 15C. The table shows the 15C at 109. But it might be that the fine detail of what should be counted makes 109 more correct than the 104 above.) RE: Calculator benchmark - Valentin Albillo - 08-17-2021 01:31 AM . Hi, EdS2: Thanks a lot for posting the information I requested, much appreciated. Having a look at it, I have some comments, read on (my highlighting) ... (08-16-2021 06:22 AM)EdS2 Wrote:(08-15-2021 03:47 PM)Valentin Albillo Wrote: What's the best solution (less keystrokes) so far for the HP-15C ? C.Ret says Total of setup = 36 keystrokes but I count 40, which coincides with the sum of the column at the far right. The TABLE ENTRY and RESULTS' keystroke count is probably wrong as well, as he says, for instance that the sequence
is 8 keystrokes long while it's actually 9, so the total count for this part might be wrong too. EdS2 Wrote:The table shows the 15C at 109. But it might be that the fine detail of what should be counted makes 109 more correct than the 104 above.) There's no "fine detail of what should be counted", it must be keystrokes for the HP-15C, and a keystroke has no ambiguity, either you press a key or you don't. Anyway, the point is moot because the 104 count is highly suspect as there are errors in both the Setup and the Computation/Results posted counts. Also, apart from the wrong count, the Setup part is less than optimal, you can significantly reduce the keystroke count as well, so it's still not clear what the minimum count for the HP-15C would be and if it that would make it the "top" model for this particular benchmark. Thanks again and best regards. V. RE: Calculator benchmark - jhallen - 08-17-2021 03:56 AM Yes, I will update when I have free time. I did update the HP-48 page- I was curious about programming the HP-48, so I've implemented the HP-15c style method on the HP-48: (should be a lot better on the HP-28..) Code:
93 keystrokes for setup. Code:
61 keystrokes for execution. RE: Calculator benchmark - C.Ret - 08-18-2021 08:20 PM HP-41C SETUP: using Alpha register to display X & Y from A & B in registers 01 & 02 respectively. Code: PACK 4 [shift][GTO][ . ][ . ] EXECUSION: each press on R/S key display X and Y in the display. Code: 3,12 STO 01 6 [ 3 ][ . ][ 1 ][ 2 ][STO][∑+ ] SCORE: 113 keystrokes RE: Calculator benchmark - jhallen - 08-19-2021 12:23 AM Sweet! I knew HP-41 could get a better score, had no idea there was enough space to show two results at once on the display. (08-18-2021 08:20 PM)C.Ret Wrote: HP-41C RE: Calculator benchmark - C.Ret - 08-19-2021 04:37 AM (08-17-2021 01:31 AM)Valentin Albillo Wrote: C.Ret says Total of setup = 36 keystrokes but I count 40, which coincides with the sum of the column at the far right. Thanks a lot Valentin for pointing out all these erroneous indications. This happened too many times since I post too fast and don't take the time to carefully read back my tippings. It's because I am so excited to have found a solution and I am distracted by translation into english (not my native language) and also by immediately looking for any alternate solution... I hope my last and further posts will contain less erroneous counts and sums since I am using an Excel Spreadsheet instead of my traditional paper and pencil scratch-pad. Thanks a lot for your attentive readings and corrections. RE: Calculator benchmark - Valentin Albillo - 08-19-2021 04:55 AM . Hi, C.Ret, (08-19-2021 04:37 AM)C.Ret Wrote:(08-17-2021 01:31 AM)Valentin Albillo Wrote: C.Ret says Total of setup = 36 keystrokes but I count 40, which coincides with the sum of the column at the far right. You're welcome. Also, did you find the fairly simple optimization of your HP-15C's Setup section which I mentioned ? You'll save a sizable number of keystrokes if you do ... Regards. V. RE: Calculator benchmark - C.Ret - 08-19-2021 05:21 PM (08-19-2021 04:55 AM)Valentin Albillo Wrote: Also, did you find the fairly simple optimization of your HP-15C's Setup section which I mentioned ? Late yesterday night, I made a few progress in this research and finally found two optimizations of my previous HP-15C code. There is a way to spare unnecessary GTO 01 and LBL 01 : First optimization: f LBL A STO 1 RCL 2 f LBL B STO 2 ... No more need of the LBL 01 label and the extra RCL 2 is no longer than the GTO 1 statement from previous version. From here, one may start the computation of X and Y. Second optimization: since the X-register always contain the value of B after the LBL B STO 2 statements, why not directly start the computation with evaluating LN(B) ? This spare a few more steps in the process ; ... g LN RCL 1 ENTER g x² + x↔y - f PSE ... to display X=A+A²-LN(B) ... g LSTx ENTER + .... to display Y=LN(B²) This two tricks reduce the original 40 keystrokes to only 36 keystrokes. But why may I stop this economy of labels so soon, is there any way to avoid any labels and the USER's mode ? Yes, like for the HP-41C, no more GTO , A or B keys, the R/S key only run all the table. This lead me to this last proposition for the HP-15C SETUP 26 keystrokes Code: g P/R f Clear PRGM 4 [ g ][P/R][ f ][Clear PRGM] Start new program EXECUSION 75 keystrokes Code: 3,12 STO 1 6 [ 3 ][ . ][ 1 ][ 2 ][STO][ 1 ] SCORE 101 PS.1: Please note that there is two f PSE instructions in this setup and the owners of the HP-15c Limited Edition (LE) will be in big troubles. The calculator display the two results X and Y successively in this order. If the user is not concentrate or fast enough to note the two results on the fly at each run, he only have to press the R/S again to repeat the displays. PS.2: The EXECUSION part is exactly the same as the one for the HP-41C. Why couldn’t the score of the venerable pocket computer be reduced to 101 ? RE: Calculator benchmark - Valentin Albillo - 08-19-2021 11:31 PM . Hi again, C.Ret, C.Ret Wrote:Valentin Albillo Wrote:Also, did you find the fairly simple [i]optimization of your HP-15C's Setup section which I mentioned ?[/i] Congratulations, this was the optimization I told you I had seen, which saves a significant number of keystrokes, namely 4, leaving the Setup section reduced from 40 keystrokes to just 36. But there's more ... C.Ret Wrote:This lead me to this last proposition for the HP-15C Once more, a teaser for you: there's again a simple optimization that will reduce your 101 score by no less than 4 keystrokes, leaving the new score as 97 ! ... See if you can find it (I'm sure you will). The HP-15C rules ! ... Regards. V. |