Hello,I found a performance issue in the definition of fit ,
src/algorithms/donut.py,
tf_session = tf.Session was repeatedly called in for col_idx in trange(len(X.columns)): and was not closed.
I think it will increase the efficiency and avoid out of memory if you close this session after using it.
Hello,I found a performance issue in the definition of
fit
, src/algorithms/donut.py, tf_session = tf.Session was repeatedly called in for col_idx in trange(len(X.columns)): and was not closed. I think it will increase the efficiency and avoid out of memory if you close this session after using it.Here are two files to support this issue,support1 and support2
Looking forward to your reply. Btw, I am very glad to create a PR to fix it if you are too busy.