Ignore:
Timestamp:
08/31/22 00:05:43 (5 weeks ago)
Author:
Maciej Komosinski
Message:

More concise code and less redundancy in dissimilarity classes, added support for archive of genotypes, added hard limit on the number of genotype chars

File:
1 edited

Legend:

Unmodified
Added
Removed
  • framspy/evolalg/dissimilarity/dissimilarity.py

    r1145 r1182  
    1111
    1212        self.output_field = output_field
    13         self.fn_reduce = None
     13        self.fn_reduce = Dissimilarity.get_reduction_by_name(reduction)
    1414        self.knn = knn
    15         if reduction == "mean": # TODO change this 'elif' sequence to dictionary?
    16             self.fn_reduce = np.mean
    17         elif reduction == "max":
    18             self.fn_reduce = np.max
    19         elif reduction == "min":
    20             self.fn_reduce = np.min
    21         elif reduction == "sum":
    22             self.fn_reduce = np.sum
    23         elif reduction == "knn_mean":
    24             self.fn_reduce = self.knn_mean
    25         elif reduction == "none" or reduction is None:
    26             self.fn_reduce = None
     15
     16
     17    @staticmethod
     18    def reduce(dissim_matrix, fn_reduce, knn):
     19        if fn_reduce is None:
     20            return dissim_matrix
     21        elif fn_reduce is Dissimilarity.knn_mean:
     22            return fn_reduce(dissim_matrix, 1, knn)
    2723        else:
    28             raise ValueError("Unknown reduction type. Supported: mean, max, min, sum, knn_mean, none")
     24            return fn_reduce(dissim_matrix, axis=1)
    2925
    30     def reduce(self, dissim_matrix):
    31         if self.fn_reduce is None:
    32             return dissim_matrix
    33         return self.fn_reduce(dissim_matrix, axis=1)
    3426
    35     def knn_mean(self, dissim_matrix,axis):
    36         return np.mean(np.partition(dissim_matrix, self.knn)[:,:self.knn],axis=axis)
     27    @staticmethod
     28    def knn_mean(dissim_matrix, axis, knn):
     29        return np.mean(np.partition(dissim_matrix, knn)[:, :knn], axis=axis)
     30
     31
     32    @staticmethod
     33    def get_reduction_by_name(reduction: str):
     34
     35        if reduction not in REDUCTION_FUNCTION:
     36            raise ValueError(f"Unknown reduction type '{reduction}'. Supported: {','.join(REDUCTION_FUNCTION.keys())}")
     37
     38        return REDUCTION_FUNCTION[reduction]
     39
     40
     41
     42REDUCTION_FUNCTION = {
     43            "mean": np.mean,
     44            "max": np.max,
     45            "min": np.min,
     46            "sum": np.sum,
     47            "knn_mean": Dissimilarity.knn_mean,
     48            "none": None
     49        }
Note: See TracChangeset for help on using the changeset viewer.