Timothy Yu

closely related to #22

related to #50; needs work due to fixed batch input size due to RNN Stateful = True Relevant: https://stackoverflow.com/questions/43702481/why-does-keras-lstm-batch-size-used-for-prediction-have-to-be-the-same-as-fittin Tensorflow can predict with different batch and step size than keras;...

  There we go: 

git lfs hooks temporarily removed + csv split into four chunks in directory `raw_data_external`: commit in `development-restructure`: dada3936d89141346834ba75cca654f42916b6e5 

in progress/start as of https://github.com/timothyyu/gdax-orderbook-ml/commit/a738775ed1e02938dd6ae017f726cc24b315ab01

autoSR_alt_experimental() start: https://github.com/timothyyu/gdax-orderbook-ml/commit/e04d0e82c91cc25de49f63b625eb45da4bc14778

e04d0e8 is start of update/rewrite of this function (python 2) to python 3: https://kite.trade/forum/discussion/1047/a-simple-python-function-to-detect-support-resistance-levels ``` def supres(ltp, n): """ This function takes a numpy array of last traded price and...

scipy argrelmin/max, scipy argrelextrema, np.greater, scipy find_peaks https://docs.scipy.org/doc/scipy/reference/generated/scipy.signal.argrelmin.html https://docs.scipy.org/doc/scipy/reference/generated/scipy.signal.argrelmax.html https://scipy.github.io/devdocs/generated/scipy.signal.find_peaks.html https://docs.scipy.org/doc/scipy/reference/generated/scipy.signal.argrelextrema.html https://docs.scipy.org/doc/numpy-1.15.1/reference/generated/numpy.greater.html

argrelextrema for local minima/maxima as of 861a2ae61cb4fc7cc0c13b032d0560e2b68349e5

autoSR() results saved to disk as of fd72791d52893d310b1c4e944e934ca83926a0f3