Use upgrade script then remediated code using v1.compatibility

make pip installable as a module

make pip installable

Import CWT package from git hub and use tensorflow 2.0 calling methods

To simplify and speed up calculations, include new option to output as 3 channels of Magnitude

Make compatible with Keras Layer so it can be used in a  Keras model, e.g., 

    tf.keras.Sequential([ComplexMorletCWT(lower_freq = 20, upper_freq = 500, n_scales = 64, stride=64, 
                                            wavelet_width = 16, fs=2042, output='Magnitude'),
                             efn.EfficientNetB4(include_top=False,weights='imagenet'),
                             L.GlobalAveragePooling2D(),
                             L.Dense(32,activation='relu'),
                             L.Dense(1, activation='sigmoid')])