Just documenting this scenario for reference / future thought...
In preparing the OME meeting workshop, I generated a Table from ROIs and wanted to filter by area ("max_points"). However, 1 image had no 'blobs' on it so the ImageJ "Analyse Particles" polygon was the whole plane.
This single outlier meant that the min/max range was huge and the histogram failed to render since it had 1950 bins!
Also the first movement of the slider instantly filtered out ALL images except the outlier.
I deleted the outlier image, but this didn't help as the Table data is loaded without checking whether images exist for each row (and looking at the code it seems this would be quite hard to do without some lengthy parsing of the data).
However, my only solution was to re-build the Table from ROIs after deleting the image, so it would be nicer if we can think of any improvement to this workaround.
Just documenting this scenario for reference / future thought...
In preparing the OME meeting workshop, I generated a Table from ROIs and wanted to filter by area ("max_points"). However, 1 image had no 'blobs' on it so the ImageJ "Analyse Particles" polygon was the whole plane. This single outlier meant that the min/max range was huge and the histogram failed to render since it had 1950 bins!
Also the first movement of the slider instantly filtered out ALL images except the outlier. I deleted the outlier image, but this didn't help as the Table data is loaded without checking whether images exist for each row (and looking at the code it seems this would be quite hard to do without some lengthy parsing of the data).
However, my only solution was to re-build the Table from ROIs after deleting the image, so it would be nicer if we can think of any improvement to this workaround.