Khicas 1.7.0 on Casio CG50, Matrices crash

[khaoskosmos]

I get crashes often using matrices, especially viewing them after a calculation. I will need to track down the exact steps to replicate it, but I'm using the latest two file version posted at Bernard's site (1.7.0) on a CG50.

https://www-fourier.univ-grenoble-alpes....sse/casio/

I wasn't sure where to post this, hopefully it will be useful to debug once I retrace the steps.

Picture:
https://imgur.com/hju8uCc

[parisse]

(07-31-2022 06:37 PM)khaoskosmos Wrote:  I get crashes often using matrices, especially viewing them after a calculation. I will need to track down the exact steps to replicate it, but I'm using the latest two file version posted at Bernard's site (1.7.0) on a CG50.

https://www-fourier.univ-grenoble-alpes....sse/casio/

I wasn't sure where to post this, hopefully it will be useful to debug once I retrace the steps.

Picture:
https://imgur.com/hju8uCc

Yes, please provide some steps to reproduce the crash. Then I can debug on the Numworks simulator or TI Nspire emulator and port the fix to the Casio.

[rprosperi]

(08-01-2022 08:55 PM)parisse Wrote:  

(07-31-2022 06:37 PM)khaoskosmos Wrote:  I get crashes often using matrices, especially viewing them after a calculation. I will need to track down the exact steps to replicate it, but I'm using the latest two file version posted at Bernard's site (1.7.0) on a CG50.

https://www-fourier.univ-grenoble-alpes....sse/casio/

I wasn't sure where to post this, hopefully it will be useful to debug once I retrace the steps.

Picture:
https://imgur.com/hju8uCc

Yes, please provide some steps to reproduce the crash. Then I can debug on the Numworks simulator or TI Nspire emulator and port the fix to the Casio.

Well, thank goodness Calculator Politics aren't totally polarized and damaged like most other politics these days! Here's a collaborative discussion about how to use a Numworks or TI tool to fix a Casio product, enabled by discussion on an HP Forum, and it's all a good thing!!

Thanks gentlemen and please do carry on....

[jonmoore]

(08-02-2022 12:13 AM)rprosperi Wrote:  

(08-01-2022 08:55 PM)parisse Wrote:  Yes, please provide some steps to reproduce the crash. Then I can debug on the Numworks simulator or TI Nspire emulator and port the fix to the Casio.

Well, thank goodness Calculator Politics aren't totally polarized and damaged like most other politics these days! Here's a collaborative discussion about how to use a Numworks or TI tool to fix a Casio product, enabled by discussion on an HP Forum, and it's all a good thing!!

Thanks gentlemen and please do carry on....

I thought the very same when reading Bernard's response.

[khaoskosmos]

Ok, I've crashed it a couple ways, but this last time was just multiplication of the 3x5 matrix by a scalar:

https://youtu.be/6wkBKjQjmBY

I had a typo entering U at one point, and I also exited the matrix editor so I had to go back. These additional steps don't make a difference to the crash.

The HP Prime is holding the CG50 up to make it relevant

Edit: not related or important, but the algb soft key alternates between centered and left aligned.

[parisse]

Thanks, but the video is unfortunately not very helpful for me. I see that you edit a 3*5 matrix named A, then you enter A in the shell, got the answer, then type 2*A, the answer is displayed in the matrixwriter, then it crashes, but I do not see which key you pressed to get the crash. I check myself a matrix creation A, then 2*A and got everything fine. It might be that your Khicas history is corrupted in some way.
I recommend that you erase your history: from Casio Memory app, erase session.xw. Then run Khicas again, try again to edit a matrix A and see if 2*A raises a crash. If the crash depends on the matrix value, instead of a video, try to report all textually. For example, tell us that A was a 3x5 matrix with value say A:=[[1,2,3,4,5],[0,0,0,0,0],[1,1,1,2,3]] and typing 2*A then pressing twice RIGHT and once UP raises a crash.

[khaoskosmos]

So far, no more crashing after erasing session.xw

I was able to multiply matrix A by a scalar and also calculated the nullspace with -trn(ker(A)), both executed as expected. I will update if I encounter the exact steps/key presses that corrupted the session.

[khaoskosmos]

I still have not tracked down the exact steps, but I think transpose is one function that causes crashes.

Using the matrix editor and and a couple matrix functions on those matrices usually causes a crash.

The last thing I did to crash was type the following, hit EXE
B:=eval(transpose(transpose([[1,2,3],[4,5,6]])))

I added the eval because noticed that the results had transpose written out instead of performed when I viewed the results. That did not help, I was able to see the results (with the function transpose), and then the calculator crashed.

[khaoskosmos]

Quick note, I was using transpose from cmds -> 3 Linear algebra -> 15 transpose(A)

This seems to be broken? I get normal results (a correct transposed matrix) when I use trn(A) but transpose(A) repeats the function into the output.

[parisse]

I have added transpose to the lexer in the alpha version of KhiCAS available here:
https://www-fourier.univ-grenoble-alpes....sio/alpha/