now using batched iteration over split string to save memory in frequ…

…ency processing

moptipy/algorithms/so/ffa/ffa_h.py

@@ -82,9 +82,9 @@ def h_to_str(
 
     >>> hl = np.array([0, 0, 1, 7, 4, 0, 0, 9, 0])
     >>> h_to_str(hl, 0)
-    '2;1;;7;;4;7;9'
+    '2;;;7;;4;7;9'
     >>> h_to_str(hl, -1)
-    '3;1;;7;;4;8;9'
+    '3;;;7;;4;8;9'
     >>> hd = Counter((1, 1, 1, 1, 1, 4, 4, 4, 4, 4, 4, 4, 3, 3, 3, 3,
     ...               3, 3, 2, 2, 2, 2, 2, 2, 2, 2, 2))
     >>> h_to_str(hd, 0)
@@ -102,7 +102,7 @@ def h_to_str(
     >>> hx[90] = 1
     >>> hx[45] = 2314
     >>> h_to_str(hx, 0)
-    '10;4;12;234;45;2314;89;111;;1'
+    '10;4;12;234;45;2314;89;111;;'
     """
     if not isinstance(h, np.ndarray | dict):
         raise type_error(h, "h", (np.ndarray, Counter))
@@ -134,7 +134,8 @@ def h_to_str(
                 write(str(use_index))
             old_index = use_index  # step the index
             write(csep)  # write separator to frequency counter
-            write(str(value))  # write the frequency
+            if value > 1:
+                write(str(value))  # write the frequency
         res: Final[str] = out.getvalue()  # get the final string
 
     if str.__len__(res) <= 0:

moptipy/evaluation/frequency.py

@@ -13,13 +13,15 @@
 import argparse
 from collections import Counter
 from gc import collect
+from itertools import batched
 from math import isfinite
 from typing import Any, Callable, Final, Iterable
 
 from pycommons.io.console import logger
 from pycommons.io.csv import CSV_SEPARATOR, SCOPE_SEPARATOR
 from pycommons.io.path import Path
 from pycommons.strings.string_conv import str_to_num
+from pycommons.strings.tools import split_str
 from pycommons.types import type_error
 
 from moptipy.algorithms.so.ffa.ffa_h import H_LOG_SECTION
@@ -460,16 +462,16 @@ def lines(self, lines: list[str]) -> bool:
             self.__state_h = 2
             counter = self.__counter
             for line in lines:
-                split = line.split(CSV_SEPARATOR)
                 prev_f: int | float = -1
-                for i in range(0, len(split), 2):
-                    cis = split[i]
-                    if str.__len__(cis) <= 0:
+                for cur_idx_s, cur_h_s in batched(map(
+                        str.strip, split_str(line, CSV_SEPARATOR)), 2):
+                    if str.__len__(cur_idx_s) <= 0:
                         cur_f = prev_f + 1
                     else:
-                        cur_f = str_to_num(cis)
+                        cur_f = str_to_num(cur_idx_s)
                     prev_f = cur_f
-                    counter[cur_f] += int(split[i + 1])
+                    counter[cur_f] += (
+                        int(cur_h_s) if str.__len__(cur_h_s) > 0 else 1)
         else:
             return super().lines(lines)
 

moptipy/version.py

@@ -2,4 +2,4 @@
 from typing import Final
 
 #: the version string of `moptipy`
-__version__: Final[str] = "0.9.124"
+__version__: Final[str] = "0.9.125"

tests/algorithms/so/ffa/test_eafea_c.py

@@ -113,5 +113,5 @@ def test_h_log() -> None:
 
         lines = tf.read_all_str().splitlines()
         assert lines[-1] == "END_H"
-        assert lines[-2] == "4;8;;9;;1"
+        assert lines[-2] == "4;8;;9;;"
         assert lines[-3] == "BEGIN_H"

tests/algorithms/so/ffa/test_fea1plus1.py

@@ -115,5 +115,5 @@ def test_h_log() -> None:
 
         lines = tf.read_all_str().splitlines()
         assert lines[-1] == "END_H"
-        assert lines[-2] == "3;2;;3;;4;;4;;2;;2;;1"
+        assert lines[-2] == "3;2;;3;;4;;4;;2;;2;;"
         assert lines[-3] == "BEGIN_H"