knime-ip / knip

KNIME Image Processing Extension
https://www.knime.com/community/image-processing
50 stars 10 forks source link

Segment Cropper synchronized after reset #395

Closed manugunkel closed 8 years ago

manugunkel commented 8 years ago

After a reset or a new loop iteration, all Segment Cropper nodes in a workflow generate identical output although they are configured to work on different segmentations. Changing the configuration once, closing the configuration and then set it back produces the right output but is unfavorable within a loop. KNIME Analytics Platform 3.2.1.v201608190927 KNIME Image Processing 1.5.1.201608031220 Win 10

dietzc commented 8 years ago

Hi @manugunkel,

thanks for reporting this problem. Can you provide an example workflow and detailed steps how we cna reproduce the problem? This would help a lot!!

Thanks,

Christian

manugunkel commented 8 years ago

Hi Christian,

the problem dissapeared. I thought it was persistent and had tested it several times, that's why I reported here. I changed the workflow to circumvent the problem. After changing back the Segment Cropper nodes show the intended and right behaviour so I'm not able to reproduce.

Sorry to bug you!!

manugunkel commented 8 years ago

Hi Christian,

ich glaug ich hab das Problem eingegrenzt, es tritt in diesem Workflow ständig auf bei allen möglichen Knoten und liegt daran, dass ich die Bilder und Label variablenbasiert umbenenne. Dann hat natürlich der entsprechende IP-Knoten nicht von vornherein den Bildnamen für die Auswahl und führt ein auto-configure image column aus - obwohl die Namen jedes mal gleich sind. Bei dieser Autokonfiguration wird dann immer das erste passende Element gewählt und der Output ist somit gleich trotz vorheriger unterschiedlicher Konfiguration.

Viele Grüße

Manuel

Am 13.09.2016 um 18:13 schrieb Christian Dietz:

Hi @manugunkel https://github.com/manugunkel,

thanks for reporting this problem. Can you provide an example workflow and detailed steps how we cna reproduce the problem? This would help a lot!!

Thanks,

Christian

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/knime-ip/knip/issues/395#issuecomment-246736052, or mute the thread https://github.com/notifications/unsubscribe-auth/AIHjWUTWh7DJoLHFjK0QVDJWD-hMVRK-ks5qpsuvgaJpZM4J6gz9.

gab1one commented 8 years ago

Hi @manugunkel, You can control the column selection using flow variables, so it should possible to always select the same column even with changing table specs. E.g using a string input node where you enter the column name.

manugunkel commented 8 years ago

Thanks Gabriel, I'll give it a shot!

Am 14.09.2016 um 11:20 schrieb Gabriel Einsdorf:

Hi @manugunkel https://github.com/manugunkel, You can control the column selection using flow variables, so it should possible to always select the same column even with changing table specs. E.g using a string input node where you enter the column name.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/knime-ip/knip/issues/395#issuecomment-246954760, or mute the thread https://github.com/notifications/unsubscribe-auth/AIHjWfRvQeceh--Nq3Sx7iIkHF985ODoks5qp7xkgaJpZM4J6gz9.

manugunkel commented 8 years ago

Perfect, solves the problem since I have the variables containing the respective names either way :)

Am 14.09.2016 um 11:20 schrieb Gabriel Einsdorf:

Hi @manugunkel https://github.com/manugunkel, You can control the column selection using flow variables, so it should possible to always select the same column even with changing table specs. E.g using a string input node where you enter the column name.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/knime-ip/knip/issues/395#issuecomment-246954760, or mute the thread https://github.com/notifications/unsubscribe-auth/AIHjWfRvQeceh--Nq3Sx7iIkHF985ODoks5qp7xkgaJpZM4J6gz9.

gab1one commented 8 years ago

I am glad I could help :smiley_cat:.