error using findThresholds: Too many input arguments.
Yaraat92 opened this issue · 15 comments
Yaraat92 commented
Hi,
I'm trying to train a classifier to detect mouse grooming behavior. I track the mice using Deeplabcut, and create the .trx structure file manually.
Everything seemed to work fine until I ran my first attempt at training, where I got the following error:
Does anyone know how I can get around this?
Yaraat92 commented
I tried to do some debugging myself and it seems that this findThresholds function i supposed to receive 2 inputs: data and params, but the line calling it from the JLabelData,m contains more arguments for some reason:
I also have another issue, I get this message when I determine the classifier features:
What does this mean?
arobie commented
Not sure about first part (Mayank will be back in the office later this
week). Can you check if perframe features calculated correctly from your
manual created trx file?
The second is the normal warning when you change window feature selections.
It will delete cached features (and the classifier trained with them) and
recompute with your new selected parameters. If you haven't changed
anything you can just hit 'cancel' instead of 'done'. Window features can
take some time to compute so we just wanted to warn users, and we didn't
want out of sync projects (windows feature selection not matching trained
classifier) to be a possible state.
Alice
…On Mon, Feb 17, 2020 at 2:22 PM Yaraat92 ***@***.***> wrote:
I tried to do some debugging myself and it seems that this findThresholds
function i supposed to receive 2 inputs: data and params, but the line
calling it from the JLabelData,m contains more arguments for some reason:
[image: תמונה3]
<https://user-images.githubusercontent.com/52376512/74681204-774d7f80-51cb-11ea-85a4-3a0431470313.png>
I also have another issue, I get this message when I determine the
classifier features:
[image: תמונה2]
<https://user-images.githubusercontent.com/52376512/74681001-ee364880-51ca-11ea-9f48-dd4023aecec4.png>
What does this mean?
—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
<#330?email_source=notifications&email_token=AAISKEEFCW6FLFSA7EYBLWLRDLPZDA5CNFSM4KWEYHOKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEL7N6DQ#issuecomment-587128590>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AAISKEEW262CKZ7KWFYZ4C3RDLPZDANCNFSM4KWEYHOA>
.
mkabra commented
You seem to be using an older version of the code. In the current version,
findThresholds accepts more than two variables as input. Can you pull the
latest version and try again?
Mayank
…On Sun, Feb 16, 2020 at 10:09 PM Yaraat92 ***@***.***> wrote:
Hi,
I'm trying to train a classifier to detect mouse grooming behavior. I
track the mice using Deeplabcut, and create the .trx structure file
manually.
Everything seemed to work fine until I ran my first attempt at training,
where I got the following error:
[image: תמונה1]
<https://urldefense.com/v3/__https://user-images.githubusercontent.com/52376512/74608625-7c3b0200-50eb-11ea-973a-512ba97b0a45.png__;!!Eh6p8Q!TvfNgs_n7oJ-bqkSstb4QolEv4OU6x_q33vtpnGXUwCnYjUwlwucUvMvCwOP9EmL_-A$>
Does anyone know how I can get around this?
—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
<https://urldefense.com/v3/__https://github.com/kristinbranson/JAABA/issues/330?email_source=notifications&email_token=AAJNKY7UPODL44CCQNC7OB3RDFTZ3A5CNFSM4KWEYHOKYY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4IN3OI5A__;!!Eh6p8Q!TvfNgs_n7oJ-bqkSstb4QolEv4OU6x_q33vtpnGXUwCnYjUwlwucUvMvCwOP8TshL7o$>,
or unsubscribe
<https://urldefense.com/v3/__https://github.com/notifications/unsubscribe-auth/AAJNKYZVGTFFE25YPCSTCCLRDFTZ3ANCNFSM4KWEYHOA__;!!Eh6p8Q!TvfNgs_n7oJ-bqkSstb4QolEv4OU6x_q33vtpnGXUwCnYjUwlwucUvMvCwOP1B9hWZU$>
.
Yaraat92 commented
Hi,
Thanks Alice for the clarification.
Mayank, thanks for your reply. I tried to re-download the current available version and ran into the same issue exactly!
Is there another source where I can get the most up-to-date version other than here?
mkabra commented
How do use JAABA code? Do you clone the JAABA git repo and update it using
pull? And on what operating system? Can you compare the findThresholds
function in your code (perframe/findThresholds.m) to the latest version
here:
https://github.com/kristinbranson/JAABA/blob/master/perframe/findThresholds.m?
As you can see, in the latest version findThresholds accepts more than two
arguments (varargin).
Mayank
…On Thu, Mar 5, 2020 at 9:38 AM Yaraat92 ***@***.***> wrote:
Hi,
Thanks Alice for the clarification.
Mayank, thanks for your reply. I tried to re-download the current
available version and ran into the same issue exactly!
Is there another source where I can get the most up-to-date version other
than here?
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<https://urldefense.com/v3/__https://github.com/kristinbranson/JAABA/issues/330?email_source=notifications&email_token=AAJNKY7SUXT3W2SJQEWRX3DRF62NDA5CNFSM4KWEYHOKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEN5PSVI*issuecomment-595261781__;Iw!!Eh6p8Q!UY3H9SjmeT_W3HBZj7xvp4SssKQzjiWP21GWKr3NIm4eGhA-Jl1XnBeuLInSFZ06zDo$>,
or unsubscribe
<https://urldefense.com/v3/__https://github.com/notifications/unsubscribe-auth/AAJNKY3OSAVC7U32MQYGFB3RF62NDANCNFSM4KWEYHOA__;!!Eh6p8Q!UY3H9SjmeT_W3HBZj7xvp4SssKQzjiWP21GWKr3NIm4eGhA-Jl1XnBeuLInSjEMNAQA$>
.
Yaraat92 commented
I use windows 64 and Matlab 2018b.
I do clone the JAABA git repo. Never updated it, but I just deleted the old version I had and cloned the current repo.
I see that in the current version on my computer the function can also receive 3 inputs but I still get the same error!
Thanks again,
Yara.
mkabra commented
Can you run "which findThresholds" and check it is the same file as you see
in the editor?
"
…On Thu, Mar 5, 2020 at 10:46 AM Yaraat92 ***@***.***> wrote:
I use windows 64 and Matlab 2018b.
I do clone the JAABA git repo. Never updated it, but I just deleted the
old version I had and cloned the current repo.
I see that in the current version on my computer the function can also
receive 3 inputs but I still get the same error!
[image: jaabaerror]
<https://urldefense.com/v3/__https://user-images.githubusercontent.com/52376512/75998285-3fec0c00-5f09-11ea-9bd2-a39c14b8203f.png__;!!Eh6p8Q!SnMxqRBXe_PikzOOlZngFB8LeF6xfYOxI8TdqlgcuiTTFk6rYDlkohdUBLVH3ovCYNs$>
Thanks again,
Yara.
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<https://urldefense.com/v3/__https://github.com/kristinbranson/JAABA/issues/330?email_source=notifications&email_token=AAJNKY7H4QJBIHHKVFSUHNTRF7CODA5CNFSM4KWEYHOKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEN5YSSY*issuecomment-595298635__;Iw!!Eh6p8Q!SnMxqRBXe_PikzOOlZngFB8LeF6xfYOxI8TdqlgcuiTTFk6rYDlkohdUBLVHcPO0xUI$>,
or unsubscribe
<https://urldefense.com/v3/__https://github.com/notifications/unsubscribe-auth/AAJNKY7D5YYNIF43W7H7QOLRF7CODANCNFSM4KWEYHOA__;!!Eh6p8Q!SnMxqRBXe_PikzOOlZngFB8LeF6xfYOxI8TdqlgcuiTTFk6rYDlkohdUBLVHtG3j44Y$>
.
Yaraat92 commented
Hi, sorry for the delay.. Things have been a bit on hold..
I did what you asked and the function is the correct one. I tried to download and run the classifier on a different computer, all from the beginning and got the same error.
What does this function do? Does it use any of the tracking information in the trx file? This is the only part of running the classifier that I am doing differently so...
mkabra commented
The function is essential for training a classifier. I'm still not certain
that the latest findThresholds.m file is being used. So you might have to
do some Matlab debugging. Can you put a breakpoint at line 5285 in
JLabelData.m and then run nargin('findThresholds') when the debugger is
stopped at that breakpoint? More info on nargin can be found at:
https://www.mathworks.com/help/matlab/ref/nargin.html
Mayank
…On Mon, Apr 6, 2020 at 7:00 PM Yaraat92 ***@***.***> wrote:
Hi, sorry for the delay.. Things have been a bit on hold..
I did what you asked and the function is the correct one. I tried to
download and run the classifier on a different computer, all from the
beginning and got the same error.
What does this function do? Does it use any of the tracking information in
the trx file? This is the only part of running the classifier that I am
doing differently so...
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<https://urldefense.com/v3/__https://github.com/kristinbranson/JAABA/issues/330*issuecomment-609795409__;Iw!!Eh6p8Q!TtSZ3EJQY3B45C_9jyUlIiPQhtyS4tJr-93yMMDiduO3D5shsGewmY8cm1S5he7PYks$>,
or unsubscribe
<https://urldefense.com/v3/__https://github.com/notifications/unsubscribe-auth/AAJNKY4WLE2BV6X7IRIRXKTRLHKO3ANCNFSM4KWEYHOA__;!!Eh6p8Q!TtSZ3EJQY3B45C_9jyUlIiPQhtyS4tJr-93yMMDiduO3D5shsGewmY8cm1S5LMNQKYA$>
.
Yaraat92 commented
I did what you suggested and the answer is '2'.
I am sure that the version is the latest because I deleted everything and downloaded the whole directory again. I also tried downloading JAABA on a different computer and had the same issue.
Could it have something to do with the warnings that appear in the attached photo?
mkabra commented
That means clearly there is another findThresholds.m function somewhere
because the one in the snapshot takes in 3 input arguments. When I had
asked you do "which" earlier, I didn't realize it'll only report a single
function. Can you try 'which -all findThresholds' (
https://www.mathworks.com/matlabcentral/answers/166740-how-do-you-resolve-conflict-when-multiple-toolboxes-have-functions-with-the-same-name),
and see how many files do you see in the output? And again might be better
to do it at the same point as you did nargin i.e., in the debugger.
Mayank
…On Thu, Apr 23, 2020 at 11:41 AM Yaraat92 ***@***.***> wrote:
I did what you suggested and the answer is '2'.
I am sure that the version is the latest because I deleted everything and
downloaded the whole directory again. I also tried downloading JAABA on a
different computer and had the same issue.
Could it have something to do with the warnings that appear in the
attached photo?
[image: warnings]
<https://urldefense.com/v3/__https://user-images.githubusercontent.com/52376512/80065136-4f1f2b80-8542-11ea-97bb-9c2fed8ecdab.png__;!!Eh6p8Q!RiFdXXMFnuejee2okf71l6qLDa3a8zmzkGfKPAHFd3u1-4Yv9pfUrcLgfmot8mxFAsE$>
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<https://urldefense.com/v3/__https://github.com/kristinbranson/JAABA/issues/330*issuecomment-618199776__;Iw!!Eh6p8Q!RiFdXXMFnuejee2okf71l6qLDa3a8zmzkGfKPAHFd3u1-4Yv9pfUrcLgfmotug_ofMo$>,
or unsubscribe
<https://urldefense.com/v3/__https://github.com/notifications/unsubscribe-auth/AAJNKY32DDDLTY2GHFQK6ADRN7LXTANCNFSM4KWEYHOA__;!!Eh6p8Q!RiFdXXMFnuejee2okf71l6qLDa3a8zmzkGfKPAHFd3u1-4Yv9pfUrcLgfmotlptT74E$>
.
Yaraat92 commented
You're apparently right...
What should I do in this case? Delete the older version ?
mkabra commented
Aah, so I think we had multiple versions of them in the repository.
Surprising that we didn't encounter this issue. I went ahead and deleted
the extra files. So if you pull the latest version there should be only one
findThresholds. Thanks for pointing it out!
Mayank
…On Thu, Apr 23, 2020 at 1:58 PM Yaraat92 ***@***.***> wrote:
You're apparently right...
What should I do in this case? Delete the older version ?
[image: findthresholds]
<https://urldefense.com/v3/__https://user-images.githubusercontent.com/52376512/80076715-3ec47c00-8555-11ea-9981-235caa1255c1.png__;!!Eh6p8Q!Ux6YDmkx-0cCQm4G2lZits8DRTGC3eXxXSbjjuT6kPFGmVsLnV-E0UvhOdaaZncWwzM$>
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<https://urldefense.com/v3/__https://github.com/kristinbranson/JAABA/issues/330*issuecomment-618259223__;Iw!!Eh6p8Q!Ux6YDmkx-0cCQm4G2lZits8DRTGC3eXxXSbjjuT6kPFGmVsLnV-E0UvhOdaa9qkszYI$>,
or unsubscribe
<https://urldefense.com/v3/__https://github.com/notifications/unsubscribe-auth/AAJNKYYKLGMYFHQDWVDPQU3RN73ZLANCNFSM4KWEYHOA__;!!Eh6p8Q!Ux6YDmkx-0cCQm4G2lZits8DRTGC3eXxXSbjjuT6kPFGmVsLnV-E0UvhOdaaGj-_UkU$>
.
Yaraat92 commented
Wow I can't believe this is solved! Thanks a lot.
Just a little question, when you say 'pull' the latest version what do you mean?
Usually I just clone and download everything from the beginning and delete the older files I have.. Is there another way?
mkabra commented
Aah, with git you don't have to manually delete and download the files. Its
been long since I used git on windows, but if you have git installed for
windows, I think right-clicking on the JAABA folder should give you an
option to pull. Once you run this it will update your local JAABA
repository to match the online one.
Mayank
…On Thu, Apr 23, 2020 at 2:10 PM Yaraat92 ***@***.***> wrote:
Wow I can't believe this is solved! Thanks a lot.
Just a little question, when you say 'pull' the latest version what do you
mean?
Usually I just clone and download everything from the beginning and delete
the older files I have.. Is there another way?
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<https://urldefense.com/v3/__https://github.com/kristinbranson/JAABA/issues/330*issuecomment-618265429__;Iw!!Eh6p8Q!WR2N_ANjDwyvGuYmRlLm_y_RUxDZkkOZqjF2miPmE6o-YWraY4v3wstyjIGTdOVDiFM$>,
or unsubscribe
<https://urldefense.com/v3/__https://github.com/notifications/unsubscribe-auth/AAJNKY3M36P5ASH3ZYANKU3RN75ITANCNFSM4KWEYHOA__;!!Eh6p8Q!WR2N_ANjDwyvGuYmRlLm_y_RUxDZkkOZqjF2miPmE6o-YWraY4v3wstyjIGThwEr75M$>
.