Rectangle select not working

[matthewspeir]

It seems like the rectangle select doesn't function after one of the latest updates.

Here's an example of a broken rectangle select, created 2019-03-11: https://genome-test.gi.ucsc.edu/~mspeir/cb_pancreas_loom/?ds=HCA_Pancreas_Aging_Loom

Here's an example with a working rectangle select, created 2019-02-26: https://genome-test.gi.ucsc.edu/~mspeir/cb_pancreas_mtx/?ds=HCA_Pancreas_mtx

[maximilianh]

No, I think this has to do with bundle_uuid being a string field with unique values. It's possible that I've never had a case like this before... I'll look into it. Usually the unique fields are the cell_id field and it's always the first one. Maybe something here is different. Thanks!

On Tue, Mar 12, 2019 at 9:52 PM Matt Speir notifications@github.com wrote:

It seems like the rectangle select doesn't function after one of the latest updates.

Here's an example of a broken rectangle select, created 2019-03-11: https://genome-test.gi.ucsc.edu/~mspeir/cb_pancreas_loom/?ds=HCA_Pancreas_Aging_Loom

Here's an example with a working rectangle select, created 2019-02-26: https://genome-test.gi.ucsc.edu/~mspeir/cb_pancreas_mtx/?ds=HCA_Pancreas_mtx

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/maximilianh/cellBrowser/issues/82, or mute the thread https://github.com/notifications/unsubscribe-auth/AAS-TSEsgPJ_L8ZLguGj05oH0jIYgW1Zks5vWBONgaJpZM4br23N .

[maximilianh]

Oh, now I understand, this makes complete sense: when you multi-select cells, the left menu tries to summarize. It tries to show histograms of the values that you selected. But with bundle_uuid this cannot work, as it's a unique identifier. I now stopped doing this.

I've made a change which I hope works, I can't test it yet, since I don't know where this data is. Can you tell me where this data is on dev?

On Wed, Mar 13, 2019 at 12:03 AM Maximilian Haeussler maximilianh@gmail.com wrote:

No, I think this has to do with bundle_uuid being a string field with unique values. It's possible that I've never had a case like this before... I'll look into it. Usually the unique fields are the cell_id field and it's always the first one. Maybe something here is different. Thanks!

On Tue, Mar 12, 2019 at 9:52 PM Matt Speir notifications@github.com wrote:

It seems like the rectangle select doesn't function after one of the latest updates.

Here's an example of a broken rectangle select, created 2019-03-11: https://genome-test.gi.ucsc.edu/~mspeir/cb_pancreas_loom/?ds=HCA_Pancreas_Aging_Loom

Here's an example with a working rectangle select, created 2019-02-26: https://genome-test.gi.ucsc.edu/~mspeir/cb_pancreas_mtx/?ds=HCA_Pancreas_mtx

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or mute the thread.

[maximilianh]

I made a release 0.4.58 with what I think is a fix for this, but haven't tested it yet.

On Wed, Mar 13, 2019 at 3:41 PM Maximilian Haeussler maximilianh@gmail.com wrote:

Oh, now I understand, this makes complete sense: when you multi-select cells, the left menu tries to summarize. It tries to show histograms of the values that you selected. But with bundle_uuid this cannot work, as it's a unique identifier. I now stopped doing this.

I've made a change which I hope works, I can't test it yet, since I don't know where this data is. Can you tell me where this data is on dev?

On Wed, Mar 13, 2019 at 12:03 AM Maximilian Haeussler maximilianh@gmail.com wrote:

No, I think this has to do with bundle_uuid being a string field with unique values. It's possible that I've never had a case like this before... I'll look into it. Usually the unique fields are the cell_id field and it's always the first one. Maybe something here is different. Thanks!

On Tue, Mar 12, 2019 at 9:52 PM Matt Speir notifications@github.com wrote:

It seems like the rectangle select doesn't function after one of the latest updates.

Here's an example of a broken rectangle select, created 2019-03-11: https://genome-test.gi.ucsc.edu/~mspeir/cb_pancreas_loom/?ds=HCA_Pancreas_Aging_Loom

Here's an example with a working rectangle select, created 2019-02-26: https://genome-test.gi.ucsc.edu/~mspeir/cb_pancreas_mtx/?ds=HCA_Pancreas_mtx

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or mute the thread.

[matthewspeir]

The starting data for https://genome-test.gi.ucsc.edu/~mspeir/cb_pancreas_loom/?ds=HCA_Pancreas_Aging_Loom is /hive/users/mspeir/cellbrowserTest/pancreas/new_metadata_test/matrix_files/Single_cell_transcriptome_analysis_of_human_pancreas.loom

I'll update my cell browser version and test this out now.

[matthewspeir]

Hmm, I think something else sprung loose. Now the cells don't even show up. Looks like it's treating all fields as 'unique id' fields:

See this one: https://genome-test.gi.ucsc.edu/~mspeir/cb_pancreas_loom/?ds=HCA_Pancreas_Aging_Loom

or this one: https://genome-test.gi.ucsc.edu/~mspeir/cb_pancreas_loom_v2/?ds=HCA_Pancreas_Aging_Loom

Data location:

/hive/users/mspeir/cellbrowserTest/pancreas/new_metadata_test/matrix_files/cbScanpyOut_pancreas_aging_loom_v2

Command run:

cbBuild -o ~/public_html/cb_pancreas_loom_v2

[maximilianh]

Yes, as I said, I couldn't really test this. I'll fix it. Thanks for finding this, unique fields are not something I've tested a lot, they're relatively rare.

On Wed, Mar 13, 2019 at 5:05 PM Matt Speir notifications@github.com wrote:

Hmm, I think something else sprung loose. Now the cells don't even show up. Looks like it's treating all fields as 'unique id' fields:

[image: cellbrowserUniqueIdField] https://user-images.githubusercontent.com/11509990/54294771-0f2ffb80-456f-11e9-9064-e8b044608675.png

See this one:

https://genome-test.gi.ucsc.edu/~mspeir/cb_pancreas_loom/?ds=HCA_Pancreas_Aging_Loom

or this one:

https://genome-test.gi.ucsc.edu/~mspeir/cb_pancreas_loom_v2/?ds=HCA_Pancreas_Aging_Loom

Data location:

/hive/users/mspeir/cellbrowserTest/pancreas/new_metadata_test/matrix_files/cbScanpyOut_pancreas_aging_loom_v2

Command run:

cbBuild -o ~/public_html/cb_pancreas_loom_v2

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/maximilianh/cellBrowser/issues/82#issuecomment-472488880, or mute the thread https://github.com/notifications/unsubscribe-auth/AAS-TdORSujO4psjm5ysTU_6hxCQVw4Vks5vWSGtgaJpZM4br23N .

[matthewspeir]

Looks like this is fixed in the latest (0.4.62). Closing