Problem with mlread

Dear Jaewon,
I used the mlread function of the latest ML version (2.2.22) to read my bhv file produced by old ML version, but it got jumbled. I used the latest ML to produce the new bhv file and it was loaded normally. Is it the structure of the bhv file changed in the new version?

Thanks,
Ming


>> mlread
An error occurred while reading the variable, 'MLConfig'.

An error occurred while reading the variable, 'on_CR.m IO struct


SignalType char
 Reward Adaptor char
 nidaq DevID char
 Dev1 Subsystem char
DigitalIO Channel double

 DIOInfo cell
  double

 char
 out Touchscreen logical


 RunMessageLoop logical

 USBJoystick char
 None
EyeTracker struct

  Name char
 None ID char ViewPoint struct


IP_address char
 169.254.110.159 Port char
 5000 Source double  @ $@ ð? ð? ð? ð? ð? ð? à? à? 4@ 4À ð? ð? ð? ð? ð? ð? EyeLink struct



It seems that your file is corrupted due to an error in the storage. I will take a look if you send me the file.

Thank you for your reply. I have sent you the bhv2 file by email. By the way, I am sure that my bhv2 file isn't corrupted because it can be normally loaded by ML edition NIMH_MonkeyLogic_2(Dec-14-2017).

Your file contains non-alphabetical characters without the information of their encoding. It is okay as long as you use the localized Windows and MATLAB, but the file may not be readable on the Windows of different languages or other operating systems, even if you try with the old version of NIMH ML. This issue was patched almost three years ago.

Download the NIMH ML package again and you will be able to read the file with the version 2.2 on your computer. However, if you want the file to be compatible with operating systems of different languages, you need to convert them. The following command will create a new BHV2 file with the encoding information in it.

convert_format('bhv2')

See the function manual for more.
https://monkeylogic.nimh.nih.gov/docs_RuntimeFunctions.html#convert_format

----------

The file also will not be readable in R2020a or later. There is still a way to read it, but just convert the file before upgrading your MATLAB.

Thank you!

I came up with another solution. I think this may be better since you don't need to convert the files if you have many of them. Type the following two lines on the MATLAB command window before calling mlread().

setpref('NIMH_MonkeyLogic','DefaultEncoding','GB18030')  % try a different encoding if yours is not 'GB18030'
setpref('NIMH_MonkeyLogic','MachineFormat','ieee-le')

-----

See this link for summary.
https://monkeylogic.nimh.nih.gov/board/read.php?2,1166,1166#msg-1166

I tried these two methods, but it still got jumbled. I've been told by senior that this file used to open with monkeylogic 2 in 2017, so I download the latest version 2.0.243, and it opened successfully, whatever old bhv file or new bhv file. The root of this bug may be the version differences between 2.0 and 2.2.

You must update v2.2 again. What you said is not possible, because the related files included in v2.0 and v2.2 are identical.

Yeah, it opened successfully after I remove ML 2.0 and reload ML 2.2. I'm astonished about it because I found the same problem on two computers of my lab. Now it runs normally. Thank you very much!