Reproducibility of results for "Measuring root mass" example

[uschille]

How could the content be improved?

I taught a hybrid workshop at FZ Juelich and participants pointed out a discrepancy in the results for the root mass in the Thresholding episode. I reckon the numbers are slightly different depending on the exact version of scikit-image, Numpy, etc., so we may want to update the numbers or include a note regarding version numbers.

(When teaching the workshop, this is actually a good opportunity to discuss reproducibility.)

For reference, these are the numbers currently on the web version:

data/trial-016.jpg,0.0482436835106383
data/trial-020.jpg,0.06346941489361702
data/trial-216.jpg,0.14073969414893617
data/trial-293.jpg,0.13607895611702128

Which part of the content does your suggestion apply to?

https://datacarpentry.github.io/image-processing/07-thresholding.html#application-measuring-root-mass