fcit
fcit copied to clipboard
kjchalup
A decision-tree based conditional independence test.
I made some minor modifications to have the option of fixing the code's randomness, allowing the test's reproducibility.
At what size dimension should I expect the length of time to be less than 5 seconds? I expected that with these dimensions it wouldn't take longer than a second....
My dataset has mixed datatypes (Continuous and DIscrete), When I try to perform conditional independence test the following error occurs: databricks/python/lib/python3.7/site-packages/fcit/fcit.py:193: RuntimeWarning: invalid value encountered in true_divide t, p_value =...
Line 37: np.random.seed(seed or int(time.time())) throws the error "TypeError: 'int' object is not callable". Consider replacing it with np.random.seed = seed or int(time.time()).
I'm just taking a look at your second example of the README. > Generate some data such that x is indpendent of y given z. > n_samples = 1000 >...
Can you please share an example how to use `fcit` CI tests with a mixture of discrete data and continuous data? I tried to one-hot encode the data, but the...
