The function calculates feature representation of each sequence from the computed attributes using the mean_val values as the attribute value for sequence. Computed attributes are stored on local file system.
Input:
config (structure): config structure
sequence (structure): A valid sequence structure.
Output:
similarity (matrix): similarity between all pairs of sequences
sequences (structure): modified structure of sequences where the sequences that does not have attributes computed are removed
feature_vector_scaled (matrix): feature vector for each sequence (row-wise). If config.hamming_features is set to 1, feature_vector_scaled is binary representation of the feature_vector_realval (clustered to 2 classes) otherwise feature_vector_scaled = feature_vector_realval
feature_vector_realval (matrix): (0,1) scaled feature vector