JCSDA-internal/eva

Dynamic option for outliers

Opened this issue Ā· 0 comments

I tested using different percentage values rather than %100 and this doesn't work. Could use some šŸŽ¶ refactoring šŸŽµ

# Decide how many values to throw out on each end of the dataset
n = datavar.size
n_throw_out = ((100-percentage_capture) * n / 100) / 2
# The value needs to be an integer
n_throw_out = np.floor(n_throw_out).astype(int)