Maybe you could patch the CPU speed to eigenmath in orther to make it work much faster, no ?
First, don't double post. Use the little edit button if its been less than 24 hours since you last posted in the same topic.
Next, Overclocking is bad for the calculator in the long run. The reason that the calculators are underclocked in the first place is to save battery power and also so that the calculations are more accurate. I think that's it.
Next, Overclocking is bad for the calculator in the long run. The reason that the calculators are underclocked in the first place is to save battery power and also so that the calculations are more accurate. I think that's it.
Quote:
PRIZM is a great calculators as, ClassPad 2, HP Prime & NSpire CX CAS
Assuming "ClassPad 2" refers to the fx-CP400 II, sorry, that's a mediocre calculator, little more than a decade-old CP300/330 with a color screen...
The firmware makes little use of color, the math abilities aren't much better than those of the CP300/330.
The slowness is annoying for some high school level simple algorithms: all of the TI-Z80, TI-68k, Nspire, Prizm, 50g, Prime series complete one particular algorithm in 3 s at worse, but the fx-CP400 needs 400s in exact mode (two orders of magnitude slower !), or something like 40s in approx mode (an order of magnitude slower).
The hardware characteristics are worse than a '2007 Nspire Clickpad.
And the worst thing in the fx-CP400, adding insult to injury, is the horrendously expensive price tag, even higher than that of a Prime or Nspire CX...
BTW, it's not because he isn't, or I am not, a moderator / administrator that you can talk everything you want to him or me
Member of the TI-Chess Team.
Co-maintainer of GCC4TI (GCC4TI online documentation), TIEmu and TILP.
Co-admin of TI-Planet.
Co-maintainer of GCC4TI (GCC4TI online documentation), TIEmu and TILP.
Co-admin of TI-Planet.
In order to compete with the TI-84 Plus C Silver Edition in terms of BASIC math speed, the ClassPad II would need to be overclocked to 4.3 GHz.
When your 200 (?) MHz calc can't even get close to the speed of a bottlenecked 15 MHz Z80 color calc during equation solving, there really is a serious problem.
When your 200 (?) MHz calc can't even get close to the speed of a bottlenecked 15 MHz Z80 color calc during equation solving, there really is a serious problem.
Ronald G V A (MD-1032) wrote:
Is there any problem with my faces, nerd??
I'm talking with a great developer as glb08ma not with you,
Wrong. This is a forum, which means the discussion is public. If you want to have a private discussion, use email or the PM system.
Ronald G V A (MD-1032) wrote:
I can talk everything that I want, you're not a moderator or the administrator. Don't be angry because nobody use your PNGViewer, you know what? It's very mediocre and it's buggy
I don't even know what you mean by your first sentence, please learn to use correct English grammar. That said, administrator here, and you actually can't say whatever pops into your head. We have both rules and guidelines about how discussions on Cemetech should proceed, and among those we suggest heavily that you not be rude.
Also, I've combined your posts. In the future, please refrain from double posting in such a short span of time.
Wow this is more activity than I have seen in one day for a while in the prizm section.
I will be responding to replies that I have something to say in reverse order (most recent first).
Ronald G V A (MD-1032): I wonder why. (That being a sarcastic remark)
DJ_O and Lionel Debroux: The slowness may be a result of poor software not poor hardware. What cpu does it have? I'd imagine the same or similar to what is in the prizm.
ordelore: You are right about battery life but there should be no difference for accuracy. The only instance I have encountered where clock speed affects accuracy is running ADCs at a higher clock speed decreased resolution.
Ronald G V A (MD-1032): I was not referring to your face which I have never seen so therefore I cannot make conclusions on that. You used lots of "emotions" in your post. Too many in my personal opinion. This is just a nitpick so lets not waste lots of time on this. Also the casio prizm has enough ram for great add-ins that I and others have made. Maybe you have not looked enough. Also calling the TI calculators "average" really depends on your perception. They are average in the sense that lots of people own them due to the fact that TI encourages schools to use their calculators. However in hardware wise the most commonly used TI calculators (83 and 84 series) are far below average in hardware. Remember when people have a monopoly there is no need to innovate. I understand that TI does not have complete monopoly but they close to having one. How else can you sell something based on a z80 and a low resolution monochrome screen? On the plus side we are free to vote with our wallets and buy better calculators. Also you insult me then put a smiley face. I do not know what to think. If my program does have a bug please send me the file that causes the bug and steps to reproduce the problem. Please post that in my thread so we do not further derail this thread. Also what do you mean by nobody? On the cemetech archives my png viewer has been downloaded 83 times as of today. 83!=nobody. Also overclocking only works on some prizms.
gbl08ma: That is quite strange. I have had issues of picojpeg working only with certain compilers. I solved my problem by switching to TJpgDec. Anyways I will try adding TJpgDec to my image viewer and we will see how well it works. The saving feature does look very good. Very nice work.
I will be responding to replies that I have something to say in reverse order (most recent first).
Ronald G V A (MD-1032): I wonder why. (That being a sarcastic remark)
DJ_O and Lionel Debroux: The slowness may be a result of poor software not poor hardware. What cpu does it have? I'd imagine the same or similar to what is in the prizm.
ordelore: You are right about battery life but there should be no difference for accuracy. The only instance I have encountered where clock speed affects accuracy is running ADCs at a higher clock speed decreased resolution.
Ronald G V A (MD-1032): I was not referring to your face which I have never seen so therefore I cannot make conclusions on that. You used lots of "emotions" in your post. Too many in my personal opinion. This is just a nitpick so lets not waste lots of time on this. Also the casio prizm has enough ram for great add-ins that I and others have made. Maybe you have not looked enough. Also calling the TI calculators "average" really depends on your perception. They are average in the sense that lots of people own them due to the fact that TI encourages schools to use their calculators. However in hardware wise the most commonly used TI calculators (83 and 84 series) are far below average in hardware. Remember when people have a monopoly there is no need to innovate. I understand that TI does not have complete monopoly but they close to having one. How else can you sell something based on a z80 and a low resolution monochrome screen? On the plus side we are free to vote with our wallets and buy better calculators. Also you insult me then put a smiley face. I do not know what to think. If my program does have a bug please send me the file that causes the bug and steps to reproduce the problem. Please post that in my thread so we do not further derail this thread. Also what do you mean by nobody? On the cemetech archives my png viewer has been downloaded 83 times as of today. 83!=nobody. Also overclocking only works on some prizms.
gbl08ma: That is quite strange. I have had issues of picojpeg working only with certain compilers. I solved my problem by switching to TJpgDec. Anyways I will try adding TJpgDec to my image viewer and we will see how well it works. The saving feature does look very good. Very nice work.
I know this is not the topic of the thread, but since the thread is mine and has gone offtopic already, it's good to see some activity (though flamebaity) in the almost-forgotten Prizm section, and these things have to be discussed somewhere, I thought I could leave some things clear, for future reference:
- Overclocking is bad, but to explain why, let me put things in another perspective: the Prizm comes clocked at 58 MHz, not just to save battery power, but also to ensure the CPU can cool itself passively, and because that's the speed CPU peripherals, for example the flash memory, expect the CPU to be running at.
In fact, some (possibly a large portion) Prizms can't be overclocked to 94.3 MHz. On those, problems like flash read errors start to be noticed at lower speeds, usually ending in a system error or a instant reboot. And the others can't go above 94.3 MHz, or 101 MHz when connected through USB, indicating possible issues with power management. The hardware was not designed to handle the power the CPU consumes when running above 58 MHz, which may damage the CPU itself and other components of the board.
This explains why you should not always have your calculator at the maximum speed, and also explains why add-ins should not overclock automatically: if they do, they won't work on the calculators where overclock doesn't work correctly; additionally, battery will be spent more quickly, leaving users wondering why their calculator batteries are already empty two weeks after swapping them, when usually the batteries lasted for a month, or more.
On the other hand, I ran my Prizm overclocked to 94.3 MHz for over a year without problems of any kind (until it broke due to provably unrelated reasons). This anecdotal evidence is proof that, when it comes to overclocking any processing unit, "your mileage may vary", and enforcing a overclocking setting on users is just plain wrong.
- While it is true some Casio calculator models suffer from extreme slowness problems, discussion about it definitely doesn't belong in this thread. However, I'd like to leave the reminder that if a model is bad, you should vote with your money...
- This Eigenmath port and the official versions of Eigenmath, including the proprietary[1] ones for Mac OS X, lack functionality which is nowadays considered standard for a CAS. Now, please carefully look and note how I never said, in this thread or anywhere else, that Eigenmath was a CAS, and the Eigenmath creator, George Weigt, never said it either as far as I know. "Eigenmath is a symbolic math program for people who need to compute with symbols as well as numbers." If it is lacking some features, it's not my fault: you are paying nothing for it, I'm not entitled to give support for it (even though I often do it, voluntarily) and it's me who decides which features go in this port. It is free software under the GNU GPL v2, so of course feel free to fork it and distribute your super version under the terms of that license, as long as you please a) give credit and b) make clear that it's not my port, but a fork of it (you should probably contact me beforehand to make sure I'm OK with the way it is presented). Same applies for Utilities.
- ProgrammerNerd is a developer as good or better than me, and in fact has helped me optimize both Utilities and Eigenmath. And taking a reasonable stance doesn't indicate friendship with anybody.
- Doesn't this forum have some rule against the excessive use of smileys? If it doesn't, it's understandable, it means it never had to deal with certain kinds of writing styles.
As a final note, I am nobody to discuss the rules and the actions of moderators and admins, but if I was a moderator here, some user(s) would be banned and some post(s) deleted or moved to another thread (my own included) already.
[1] or perhaps no longer, since there appears to be an updated source ZIP with the "shell-mode" edition here: http://gweigt.net/src.zip . The license is unknown. I should probably update the About screen of my port.
- Overclocking is bad, but to explain why, let me put things in another perspective: the Prizm comes clocked at 58 MHz, not just to save battery power, but also to ensure the CPU can cool itself passively, and because that's the speed CPU peripherals, for example the flash memory, expect the CPU to be running at.
In fact, some (possibly a large portion) Prizms can't be overclocked to 94.3 MHz. On those, problems like flash read errors start to be noticed at lower speeds, usually ending in a system error or a instant reboot. And the others can't go above 94.3 MHz, or 101 MHz when connected through USB, indicating possible issues with power management. The hardware was not designed to handle the power the CPU consumes when running above 58 MHz, which may damage the CPU itself and other components of the board.
This explains why you should not always have your calculator at the maximum speed, and also explains why add-ins should not overclock automatically: if they do, they won't work on the calculators where overclock doesn't work correctly; additionally, battery will be spent more quickly, leaving users wondering why their calculator batteries are already empty two weeks after swapping them, when usually the batteries lasted for a month, or more.
On the other hand, I ran my Prizm overclocked to 94.3 MHz for over a year without problems of any kind (until it broke due to provably unrelated reasons). This anecdotal evidence is proof that, when it comes to overclocking any processing unit, "your mileage may vary", and enforcing a overclocking setting on users is just plain wrong.
- While it is true some Casio calculator models suffer from extreme slowness problems, discussion about it definitely doesn't belong in this thread. However, I'd like to leave the reminder that if a model is bad, you should vote with your money...
- This Eigenmath port and the official versions of Eigenmath, including the proprietary[1] ones for Mac OS X, lack functionality which is nowadays considered standard for a CAS. Now, please carefully look and note how I never said, in this thread or anywhere else, that Eigenmath was a CAS, and the Eigenmath creator, George Weigt, never said it either as far as I know. "Eigenmath is a symbolic math program for people who need to compute with symbols as well as numbers." If it is lacking some features, it's not my fault: you are paying nothing for it, I'm not entitled to give support for it (even though I often do it, voluntarily) and it's me who decides which features go in this port. It is free software under the GNU GPL v2, so of course feel free to fork it and distribute your super version under the terms of that license, as long as you please a) give credit and b) make clear that it's not my port, but a fork of it (you should probably contact me beforehand to make sure I'm OK with the way it is presented). Same applies for Utilities.
- ProgrammerNerd is a developer as good or better than me, and in fact has helped me optimize both Utilities and Eigenmath. And taking a reasonable stance doesn't indicate friendship with anybody.
- Doesn't this forum have some rule against the excessive use of smileys? If it doesn't, it's understandable, it means it never had to deal with certain kinds of writing styles.
As a final note, I am nobody to discuss the rules and the actions of moderators and admins, but if I was a moderator here, some user(s) would be banned and some post(s) deleted or moved to another thread (my own included) already.
[1] or perhaps no longer, since there appears to be an updated source ZIP with the "shell-mode" edition here: http://gweigt.net/src.zip . The license is unknown. I should probably update the About screen of my port.
Hey, can you tell me how many seconds does eigenmath take (in average) to solve a derivative or an integral ? Does it give the result right away or can it take over 10seconds for a complicated function ?
How much new fonctionnalities from SIMB do you think you could add without making eigenmath slow ? Mabe you couldadd somevery important fonctionnalities to eigenmath for the 1.0 version
Thanks and congratulations !! what you did from the first version to beta8 is just great !!
How much new fonctionnalities from SIMB do you think you could add without making eigenmath slow ? Mabe you couldadd somevery important fonctionnalities to eigenmath for the 1.0 version
Thanks and congratulations !! what you did from the first version to beta8 is just great !!
jpto wrote:
Hey, can you tell me how many seconds does eigenmath take (in average) to solve a derivative or an integral ? Does it give the result right away or can it take over 10seconds for a complicated function ?
It varies, why don't you see for yourself?
Quote:
How much new fonctionnalities from SIMB do you think you could add without making eigenmath slow ? Mabe you couldadd somevery important fonctionnalities to eigenmath for the 1.0 version
As I said it's very hard to port things from SMIB because it is based on an earlier version of Eigenmath and its functions are usually much more resource-intensive. Adding complex things from it requires that dependencies be added too, which increases the g3a size (but that's not the main problem) and reduces the amount of RAM available for variables, command execution and display (and that's the main problem).
Also, a large portion of SMIB is based on built-in scripting, which can't be used on this port due to the small symbol table (you'll notice that the more you put on the startup script, for example, the less memory you have for operations).
Too whoever created this port!
My friend and i who both have the cg20 just found out that there are no restrictions for using Cas type systems for calculators on our final in Calculus 1.
We both tried to find some cheap ti89's, but we decided to take one quick look if sombody had made something like this program for the cg20.
Short story short, we found this forum and the dl link for the software. Now we both have Eigenmath, and it is awesome!!
Thank you!!!
My friend and i who both have the cg20 just found out that there are no restrictions for using Cas type systems for calculators on our final in Calculus 1.
We both tried to find some cheap ti89's, but we decided to take one quick look if sombody had made something like this program for the cg20.
Short story short, we found this forum and the dl link for the software. Now we both have Eigenmath, and it is awesome!!
Thank you!!!
Anyone know how you kan find the determinant for a matrix in Eigenmath? I googled it but I didn't get it to work. :/
Use the "det" command.
From the Eigenmath manual:
From the Eigenmath manual:
Quote:
det
det(m) returns the determinant of matrix m.
det(m) returns the determinant of matrix m.
gbl08ma wrote:
Use the "det" command.
From the Eigenmath manual:
From the Eigenmath manual:
Quote:
det
det(m) returns the determinant of matrix m.
det(m) returns the determinant of matrix m.
Yepp and according to the site I was looking on it said to write your matrix like ((1,2,3),(3,4,5),(6,7,)
So I tried "det((1,2,3),(3,4,5),(6,7,) but it didn't work... but I guess I am doing it wrong?
I imagine you'd have to write "det(((1,2,3),(3,4,5),(6,7,8 )))" because if m is ((1,2,3),(3,4,5),(6,7,), det(m) will have one more couple of parenthesis. I dont have my calculator to try, it's just a guess.
Nemhardy wrote:
I imagine you'd have to write "det(((1,2,3),(3,4,5),(6,7,8 )))" because if m is ((1,2,3),(3,4,5),(6,7,), det(m) will have one more couple of parenthesis. I dont have my calculator to try, it's just a guess.
det(((1,2,3),(3,4,5),(6,7,8 ))) worked, det((1,2,3),(3,4,5),(6,7,8 )) returned ((1,2,3))...
Thanks for the help!
Version 1.0 (stable) of Eigenmath is finally ready and shipping!
Changes since beta 8:
- Fixed reboot/system error when trying to copy the last result at a point when Eigenmath was yet to run;
- Added optional session persistence (console messages, command history, and variables are automatically saved when exiting, and restored when opening)
- Command history entries doubled from 20 to 40, and maximum command length reduced from 1000 characters to a sensible 500.
- Greek letter names (alpha, beta, pi, theta, etc.) are shown as actual Greek characters;
- Roots (anything with exponent in the form 1/x) are pretty-printed, along with brackets;
- Broken limit solver removed;
- Usage of a specific folder in storage memory (@EIGEN) for all Eigenmath data files, including the startup script;
- Better text input (for example, the cursor is no longer locked at the right of the screen when editing long commands);
- Support for partial command input (press Shift+EXE);
- Different memory management, basically there are two 128 KB heaps now (the OS one plus another on the stack), each is used for different kinds of data. Practical takeout: there is now space for more symbols and more complex contents on those symbols, without taking out space for computations with them.
- Fixed bug where graph mode would not be completely left when the user exited it;
- Fixed excessive memory usage/possible overflow when recording a script;
- Fixed nroots function (broken since at least Beta 8, possibly since Beta 5 or earlier, had to do with compiler (mis)optimizations so I'm not sure if the older binaries had the problem);
- The current input is no longer cleared when pressing a custom key (and the handler is on), or when loading a script;
- Many other bug fixes, optimizations and cosmetic changes.
Usage instructions: see https://github.com/gbl08ma/eigenmath/wiki/Usage-instructions (lots of screenshots!)
Migrating from previous versions: apart from replacing the g3a, if you have a startup script, you need to create a folder called @EIGEN on the root of the storage memory (that is, out of any folders) and move the startup script there.
The @EIGEN folder will also contain two files needed for session persistence. To disable session persistence (doesn't automatically delete the storage memory files), press Shift then Menu and change the "Save Session" setting.
Suggested startup script: see here.
Download
...at the usual URL: http://tny.im/prEigenDL
(which doesn't include the startup script or a readme, it's just the g3a);
...the same thing, but packaged in a nice ZIP with readme, license and startup script: http://tny.im/prEigenZipDL
or soon, from your favorite calculator archives.
As I explained before, since this is a stable release, feel free to redistribute it - but, if uploading e.g. to file archives, please redistribute the ZIP and not just the g3a. For transferring from one calc to another, sending just the g3a is fine, of course.
Why you should update
If you're still using a really old version, like the first Betas, it will look like a massive upgrade to your calculator. Especially, if you are still using the Beta 1 (which certain sites were still redistributing until recently), it will look like a upgrade from a almost-falling-apart-bike to a solid supersonic jet (make sure to read the usage instructions so you don't miss any feature).
If you're using a more recent version, you'll get new, exciting features like more advanced pretty-printing, text input that doesn't suck, optional session saving, more stability and bug fixes, and much more.
Updating is not hard and takes less than 10 minutes; plus, by connecting your Prizm to the computer, you'll get the opportunity to backup your documents, reorganize them easily, and add more apps and games.
---
As always, feel free to ask any questions and report any problems.
Changes since beta 8:
- Fixed reboot/system error when trying to copy the last result at a point when Eigenmath was yet to run;
- Added optional session persistence (console messages, command history, and variables are automatically saved when exiting, and restored when opening)
- Command history entries doubled from 20 to 40, and maximum command length reduced from 1000 characters to a sensible 500.
- Greek letter names (alpha, beta, pi, theta, etc.) are shown as actual Greek characters;
- Roots (anything with exponent in the form 1/x) are pretty-printed, along with brackets;
- Broken limit solver removed;
- Usage of a specific folder in storage memory (@EIGEN) for all Eigenmath data files, including the startup script;
- Better text input (for example, the cursor is no longer locked at the right of the screen when editing long commands);
- Support for partial command input (press Shift+EXE);
- Different memory management, basically there are two 128 KB heaps now (the OS one plus another on the stack), each is used for different kinds of data. Practical takeout: there is now space for more symbols and more complex contents on those symbols, without taking out space for computations with them.
- Fixed bug where graph mode would not be completely left when the user exited it;
- Fixed excessive memory usage/possible overflow when recording a script;
- Fixed nroots function (broken since at least Beta 8, possibly since Beta 5 or earlier, had to do with compiler (mis)optimizations so I'm not sure if the older binaries had the problem);
- The current input is no longer cleared when pressing a custom key (and the handler is on), or when loading a script;
- Many other bug fixes, optimizations and cosmetic changes.
Usage instructions: see https://github.com/gbl08ma/eigenmath/wiki/Usage-instructions (lots of screenshots!)
Migrating from previous versions: apart from replacing the g3a, if you have a startup script, you need to create a folder called @EIGEN on the root of the storage memory (that is, out of any folders) and move the startup script there.
The @EIGEN folder will also contain two files needed for session persistence. To disable session persistence (doesn't automatically delete the storage memory files), press Shift then Menu and change the "Save Session" setting.
Suggested startup script: see here.
Download
...at the usual URL: http://tny.im/prEigenDL
(which doesn't include the startup script or a readme, it's just the g3a);
...the same thing, but packaged in a nice ZIP with readme, license and startup script: http://tny.im/prEigenZipDL
or soon, from your favorite calculator archives.
As I explained before, since this is a stable release, feel free to redistribute it - but, if uploading e.g. to file archives, please redistribute the ZIP and not just the g3a. For transferring from one calc to another, sending just the g3a is fine, of course.
Why you should update
If you're still using a really old version, like the first Betas, it will look like a massive upgrade to your calculator. Especially, if you are still using the Beta 1 (which certain sites were still redistributing until recently), it will look like a upgrade from a almost-falling-apart-bike to a solid supersonic jet (make sure to read the usage instructions so you don't miss any feature).
If you're using a more recent version, you'll get new, exciting features like more advanced pretty-printing, text input that doesn't suck, optional session saving, more stability and bug fixes, and much more.
Updating is not hard and takes less than 10 minutes; plus, by connecting your Prizm to the computer, you'll get the opportunity to backup your documents, reorganize them easily, and add more apps and games.
---
As always, feel free to ask any questions and report any problems.
Register to Join the Conversation
Have your own thoughts to add to this or any other topic? Want to ask a question, offer a suggestion, share your own programs and projects, upload a file to the file archives, get help with calculator and computer programming, or simply chat with like-minded coders and tech and calculator enthusiasts via the site-wide AJAX SAX widget? Registration for a free Cemetech account only takes a minute.
» Go to Registration page
» Go to Registration page
Page 7 of 9
» All times are UTC - 5 Hours
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
Advertisement