Changeset 971

Timestamp:

06/30/20 14:42:27 (4 years ago)

Author:

Maciej Komosinski

Message:

Added support for calculation of dissimilarity of more than two genotypes at once (returns a full square matrix)

File:

• framspy/FramsticksCLI.py (modified) (3 diffs)

framspy/FramsticksCLI.py

@@ +1 @@
+from math import sqrt
 from subprocess import Popen, PIPE, check_output
 from enum import Enum
-from typing import List
+from typing import List  # to be able to specify a type hint of list(something)
 from itertools import count  # for tracking multiple instances
 import json
@@ -216 +217 @@
 
 
-        def dissimilarity(self, genotype1: str, genotype2: str) -> float:
-                files = self.__runCommand(self.DISSIMIL_CMD, [genotype1, genotype2], self.DISSIMIL_FILE, self.GENO_SAVE_FILE_FORMAT["NATIVEFRAMS"])
+        def dissimilarity(self, genotype_list: List[str]) -> np.ndarray:
+                """
+                Returns:
+                        A square array with dissimilarities of each pair of genotypes.
+                """
+                files = self.__runCommand(self.DISSIMIL_CMD, genotype_list, self.DISSIMIL_FILE, self.GENO_SAVE_FILE_FORMAT["NATIVEFRAMS"])
                 with open(files[-1]) as f:
                         dissimilarity_matrix = np.genfromtxt(f, dtype=np.float64, comments='#', encoding=None, delimiter='\t')
                 # We would like to skip column #1 while reading and read everything else, but... https://stackoverflow.com/questions/36091686/exclude-columns-from-genfromtxt-with-numpy
                 # This would be too complicated, so strings (names) in column #1 become NaN as floats (unless they accidentally are valid numbers) - not great, not terrible
-                EXPECTED_SHAPE = (2, 4)
-                assert dissimilarity_matrix.shape == EXPECTED_SHAPE, f"Not a correct dissimilarity matrix, expected {EXPECTED_SHAPE} "
-                for i in range(len(dissimilarity_matrix)):
-                        assert dissimilarity_matrix[i][i + 2] == 0, "Not a correct dissimilarity matrix, diagonal expected to be 0"
-                assert dissimilarity_matrix[0][3] == dissimilarity_matrix[1][2], "Probably not a correct dissimilarity matrix, expecting symmetry, verify this"
-                self.__cleanUpCommandResults(files)
-                return dissimilarity_matrix[0][3]
+                square_matrix = dissimilarity_matrix[:, 2:]  # get rid of two first columns (fitness and name)
+                EXPECTED_SHAPE = (len(genotype_list), len(genotype_list))
+                # print(square_matrix)
+                assert square_matrix.shape == EXPECTED_SHAPE, f"Not a correct dissimilarity matrix, expected {EXPECTED_SHAPE} "
+                for i in range(len(square_matrix)):
+                        assert square_matrix[i][i] == 0, "Not a correct dissimilarity matrix, diagonal expected to be 0"
+                assert (square_matrix == square_matrix.T).all(), "Probably not a correct dissimilarity matrix, expecting symmetry, verify this"
+                self.__cleanUpCommandResults(files)
+                return square_matrix
 
 
@@ -281 +288 @@
         offspring = framsCLI.crossOver(parent1, parent2)
         print("\tCrossover (Offspring):", offspring)
-        print('\tDissimilarity of Parent1 and Offspring:', framsCLI.dissimilarity(offspring, parent1))
+        print('\tDissimilarity of Parent1 and Offspring:', framsCLI.dissimilarity([parent1, offspring])[0, 1])
         print('\tPerformance of Offspring:', framsCLI.evaluate(offspring))
         print('\tValidity of Offspring:', framsCLI.isValid(offspring))