Kevin Mader

yea, it won't build the layer. You can make it but as soon as you add it to a model (sequential) or call it with another layer it throws the...

an interactive plot: https://www.kaggle.com/kmader/how-k-means-works

https://www.kaggle.com/kmader/qbi-illumination-correction/

you have to reshape or flatten the array to use it with K-means. It (and most scikit-learn functions) take only 2D input. The first dimension is the number of observations...

https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/HVBW0Q

from @bollwyvl's post on the failed PR > geckodriver must be downloaded into path separately If you're on the conda bus, you still have to install `firefox` and `chrome` yourself,...

@betatim I've added the basic functionality, but the bounding boxes do not show up for some reason. The demo.ipynb is probably the best starting point for experimenting

## freehand (polygons) - get ``` freehand_shapes = cs_view.cur_image_view.get_tool_state().\ get('imageIdToolState', {}).\ get('', {}).\ get('freehand', {}).\ get('data', []) ``` setters are much more difficult if the old state is to be...

@betatim now the bounding boxes show up, but it is still a bit finicky

probably some useful ideas to be copied from the ITK widget: https://github.com/InsightSoftwareConsortium/itk-jupyter-widgets/blob/master/js/lib/viewer.js