search

kurokida / jsQUEST

A Bayesian adaptive psychometric method for measuring thresholds in online experiments.
MIT License
17 stars 4 forks source link

An intensity that was in range was reported as out of range #13

Closed tpronk closed 2 years ago

tpronk commented 3 years ago

Denis Pelli has been running tests of the PsychoJS integration of jsQUEST and got an unexpected error message:

QuestUpdate: intensity -2.1941879156398305 out of range
-2.1989700043360187 to 2.8010299956639813. Pdf will be inexact.
Suggest that you increase “range” in call to QuestCreate.
kurokida commented 3 years ago

Thank you for contacting me. I have replied in the topic.

I have an additional report.

When I run the test/jsQUEST_test.js using v1.01 which includes the bug, I received the following output.

The results of jsQUEST, MATLAB, and PsychoPy should be almost equal. jsQUEST: Final threshold estimate (mean+-sd) is 79.04674452884068 +- 4.182374537623455 MATLAB: Final threshold estimate (mean+-sd) is 79.073768 +- 4.173877 PsychoPy: Final threshold estimate (mean+-sd) is 79.33662751283129 +- 4.233607966057011

After fixing the bug, I received the following output.

The results of jsQUEST, MATLAB, and PsychoPy should be almost equal. jsQUEST: Final threshold estimate (mean+-sd) is 79.07376848839164 +- 4.173877227129627 MATLAB: Final threshold estimate (mean+-sd) is 79.073768 +- 4.173877 PsychoPy: Final threshold estimate (mean+-sd) is 79.33662751283129 +- 4.233607966057011

The results of jsQUEST and MATLAB are now even closer together.

tpronk commented 3 years ago

Nice! Here is an idea on version numbers. Maybe consider adopting semver? That way the version number communicates whether a new release is a bug fix, adds a new feature,or is a breaking change in the API. See https://devhints.io/semver

On Sat, 13 Nov 2021, 12:00 Daiichiro Kuroki, @.***> wrote:

Thank you for contacting me. I have replied in the topic https://discourse.psychopy.org/t/incoherent-error-message-from-questhandler-in-psychojs/25864 .

I have an additional report.

When I run the test/jsQUEST_test.js using v1.01 which includes the bug, I received the following output.

The results of jsQUEST, MATLAB, and PsychoPy should be almost equal. jsQUEST: Final threshold estimate (mean+-sd) is 79.04674452884068 +- 4.182374537623455 MATLAB: Final threshold estimate (mean+-sd) is 79.073768 +- 4.173877 PsychoPy: Final threshold estimate (mean+-sd) is 79.33662751283129 +- 4.233607966057011

After fixing the bug, I received the following output.

The results of jsQUEST, MATLAB, and PsychoPy should be almost equal. jsQUEST: Final threshold estimate (mean+-sd) is 79.07376848839164 +- 4.173877227129627 MATLAB: Final threshold estimate (mean+-sd) is 79.073768 +- 4.173877 PsychoPy: Final threshold estimate (mean+-sd) is 79.33662751283129 +- 4.233607966057011

The results of jsQUEST and MATLAB are now even closer together.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/kurokida/jsQUEST/issues/13#issuecomment-968057146, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACESG4AVPJ7WYIUXAEKR3MLULZHOHANCNFSM5HXZCGRA . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

kurokida commented 3 years ago

Thank you for your advice.

Let me confirm. Does it mean that I should have released 1.0.2 instead of 1.1?

I agree with it. I hadn't really thought about this. Since I have already released version 1.1, how about we follow semver from the next version?

tpronk commented 3 years ago

Yeah, the bugfix would increase the last number of the version. And agreed, let's follow this convention for future updates

On Mon, 15 Nov 2021, 00:44 Daiichiro Kuroki, @.***> wrote:

Thank you for your advice.

Let me confirm. Does it mean that I should have released 1.0.2 instead of 1.1?

I agree with it. I hadn't really thought about this. Since I have already released version 1.1, how about we follow semver from the next version?

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/kurokida/jsQUEST/issues/13#issuecomment-968407163, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACESG4A6NOOQUOJZZDQFH2TUMBJYJANCNFSM5HXZCGRA . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

kurokida commented 2 years ago

Thank you for your reply. I understand.