Joost van Griethuysen

Not necessarily. Review the changelog in the pyradiomics documentation. Amy changes affecting calculated results are detailed there.

Using the branch from #285, I was able to build and launch using MyBinder. This required some changes to the dockerfile, as 3 notebooks are run as part of the...

This is the [link](https://beta.mybinder.org/v2/gh/JoostJM/pyradiomics/fix-urllib-python3-compatibility) to the image.

[link](https://beta.mybinder.org/v2/gh/radiomics/pyradiomics/master) to the image build upon radiomics/pyradiomics/master

@mattwarkentin It's not really a computational issue, just an edge case which remains valid. The arbitrary value was suggested in the original source describing the use of NGTDM. An alteranative...

I assume you're using the featureextractor? In the `execute()` function, you can pass the value of the label. Best way to extract then would be: ``` extractor = featureextractor.RadiomicsFeatureExtractor(params) results...

Can you check your radiomics.imageoperations file (located in site packages). [This function should exist](https://github.com/AIM-Harvard/pyradiomics/blob/v3.0.1/radiomics/imageoperations.py#L1081), maybe you changed something in imageoperations?

@fbilly945, This error is precisely what is stated: you selected label 1 as your ROI to extract features. However, in the mask file you supplied, no voxels have the value...

You correctly fixed your first error. The second one occurs a bit later. Reading your stacktrace, it looks like you implemented the firstorder featureclass directly in your own custom code....

Please add the additional information as requested by this template so we are better able to help you