Fixes a bug in the initialization of random mutation (#66)

* Fixes a bug in the initialization of random mutation

* Removes test due to space requirements

* removes the test for real

* Adds the ability to provide a tokenizer in random mutation solver

* Bumps dev version

Author: miguelgondu

pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "poli-baselines"
-version = "1.0.2"
+version = "1.0.2.dev1"
 description = "poli-baselines, a library of discrete objective optimizers"
 readme = "README.md"
 authors = [{name="Miguel González-Duque", email="[email protected]"}]
@@ -94,7 +94,7 @@ markers = [
 profile = "black"
 
 [tool.bumpversion]
-current_version = "1.0.2"
+current_version = "1.0.2.dev1"
 parse = """(?x)
     (?P<major>0|[1-9]\\d*)\\.
     (?P<minor>0|[1-9]\\d*)\\.

setup.cfg

@@ -1,6 +1,6 @@
 [metadata]
 name = poli_baselines
-version = 1.0.2
+version = 1.0.2.dev1
 author_email = [email protected]
 description = Baselines for Discrete Sequence Optimization, focusing on proteins
 long_description = file: readme.md

src/poli_baselines/__init__.py

@@ -1 +1 @@
-__version__ = "1.0.2"
+__version__ = "1.0.2.dev1"

src/poli_baselines/solvers/simple/random_mutation.py

@@ -10,6 +10,8 @@
 """
 
 import random
+import warnings
+from typing import Callable
 
 import numpy as np
 from poli.core.abstract_black_box import AbstractBlackBox
@@ -28,7 +30,23 @@ def __init__(
         batch_size: int = 1,
         greedy: bool = True,
         alphabet: list[str] | None = None,
+        tokenizer: Callable[[str], list[str]] | None = None,
     ):
+        if x0.ndim == 1:
+            if tokenizer is None:
+                warnings.warn(
+                    "In the initialization of the RandomMutation solver: \n"
+                    "The input is a 1D array, but no tokenizer was provided.\n"
+                    "Assuming that the input is a string that can be tokenized\n"
+                    "character by character using list(x_i)."
+                )
+
+                def tokenizer(x):
+                    return list(x)
+
+            x0_ = [tokenizer(x_i) for x_i in x0]
+            x0 = np.array(x0_)
+
         super().__init__(black_box, x0, y0)
         self.alphabet = black_box.info.alphabet if alphabet is None else alphabet
         self.alphabet_without_empty = [s for s in self.alphabet if s != ""]

src/poli_baselines/tests/solvers/simple/test_random_mutation.py

@@ -41,7 +41,3 @@ def test_random_mutation_in_docs():
     solver.solve(max_iter=100)
 
     print(solver.get_best_solution())
-
-
-if __name__ == "__main__":
-    test_random_mutation_in_docs()