Changeset 1313 for cpp/frams/genetics/genooperators.cpp

Timestamp:

07/11/24 17:15:51 (4 months ago)

Author:

Maciej Komosinski

Message:

Color mutations in f1 and f4, and a new syntax for "allowed modifiers" (opposite to previous "excluded modifiers") with optional probabilities for each modifier

File:

• cpp/frams/genetics/genooperators.cpp (modified) (5 diffs)

cpp/frams/genetics/genooperators.cpp

@@ -1 +1 @@
 // This file is a part of Framsticks SDK.  http://www.framsticks.com/
-// Copyright (C) 1999-2023  Maciej Komosinski and Szymon Ulatowski.
+// Copyright (C) 1999-2024  Maciej Komosinski and Szymon Ulatowski.
 // See LICENSE.txt for details.
 
@@ -56 +56 @@
         for (sum = 0, i = 0; i < count; i++) { sum += probtab[i]; if (sel < sum) return i; }
         return -1;
+}
+
+int GenoOperators::roulette(const vector<double> &probtab)
+{
+        return roulette(probtab.data(), (int)probtab.size());
 }
 
@@ -461 +466 @@
 }
 
-char *GenoOperators::strchrn0(const char *str, char ch)
+char *GenoOperators::strchr_no0(const char *str, char ch)
 {
         return ch == 0 ? NULL : strchr((char *)str, ch);
 }
 
-int GenoOperators::getRandomChar(const char *choices, const char *excluded)
-{
-        int allowed_count = 0;
-        for (size_t i = 0; i < strlen(choices); i++) if (!strchrn0(excluded, choices[i])) allowed_count++;
-        if (allowed_count == 0) return -1; //no char is allowed
-        int rnd_index = rndUint(allowed_count) + 1;
-        allowed_count = 0;
-        for (size_t i = 0; i < strlen(choices); i++)
-        {
-                if (!strchrn0(excluded, choices[i])) allowed_count++;
-                if (allowed_count == rnd_index) return int(i);
-        }
-        return -1; //never happens
+double GenoOperators::probOfModifier(const char* mod_def)
+{
+        if (*(mod_def + 1) == '(') //the special syntax with the appended probability value in (...)
+                return std::atof(mod_def + 2); //0.0 when no valid number
+        return 1.0;
+}
+
+char GenoOperators::getRandomModifier(const char *choices)
+{
+        static const char* EXTRA_CHARS = "().0123456789";
+        // this function assumes that EXTRA_CHARS are only used for the special probabilities syntax in "choices", not as valid choice characters.
+        vector<char> allowed; //this could be determined only once for a given "choices", as long as the effect of "choices" is deterministic (i.e., "choices" does not include probabilities)
+        size_t choices_len = strlen(choices);
+        allowed.reserve(choices_len); //max size, avoid reallocations later
+        for (size_t i = 0; i < choices_len; i++)
+        {
+                if (strchr(EXTRA_CHARS, choices[i])) continue; //skip parentheses and numbers
+                double prob = probOfModifier(&choices[i]);
+                if (prob == 1.0 || rndDouble(1) < prob)
+                        allowed.push_back(choices[i]);
+        }
+        if (allowed.size() == 0) return 0; //no char is allowed
+        return allowed[rndUint(allowed.size())];
+}
+
+char GenoOperators::getRandomColorModifier(const char *choices, const char *color_modifiers)
+{
+        vector<char> allowed_colors;
+        vector<double> allowed_probs;
+        size_t colors_len = strlen(color_modifiers);
+        allowed_colors.reserve(colors_len); //max size, avoid reallocations later
+        allowed_probs.reserve(colors_len); //max size, avoid reallocations later
+        for (size_t i = 0; i < colors_len; i++) //for all known color modifiers...
+        {
+                const char *pos = strchr(choices, color_modifiers[i]); //...search in "choices" - i.e., in currently set active modifiers. Note that "choices" may use an extended syntax with numbers and parentheses, such as qM(0.1)Dm(0.1)dG(0.2)C
+                if (pos) //found the color modifier in choices
+                {
+                        allowed_colors.push_back(*pos);
+                        allowed_probs.push_back(probOfModifier(pos));
+                }
+        }
+        // the above "parsing" part could be done only once "choices" changes, not every time we want to get a random color modifier...
+        int idx = roulette(allowed_probs);
+        return idx < 0 ? 0 : allowed_colors[idx];
 }
 
@@ -544 +580 @@
 }
 
-string GenoOperators::simplifiedModifiers(const string & original)
+string GenoOperators::simplifiedModifiers(const string & original, const char* colorgenes)
 {
         const int MAX_NUMBER_SAME_TYPE = 5; // max. number of modifiers of each type (case-insensitive). The more characters, the closer we can get to min and max values of a given property at the expense of the length of evolved genotypes. 5 is "close enough", but how close we get to the extreme also depends on the initial value of a given property, which is not always exactly in the middle of min and max. rR is treated separately in simplification because their influence follows different (i.e., simple additive) logic - so the simplifiedModifiersFixedOrder() logic with cancelling out antagonistic modifiers would be appropriate for rR.
+        const int MAX_NUMBER_SAME_TYPE_COLOR = 1; //color does not affect fitness and is used purely for aesthetics, so allow at most 1 char for each r,g,b channel - we get very low resolution of colors (only 3*3*3 combinations), but we spare the genotype length and limit bloat
         int counter[256] = {}; //initialize with zeros; 256 is unnecessarily too big and redundant, but enables very fast access (indexed directly by the ascii code)
         string simplified = "";
@@ -556 +593 @@
                 unsigned char lower = std::tolower(c);
                 counter[lower]++;
-                if (counter[lower] <= MAX_NUMBER_SAME_TYPE) //get rid of modifiers that are too numerous, but get rid of the first ones in the string (="oldest", the last ones looking from the end), because their influence on the parameter value is the smallest
+                int MAX_NUMBER = strchr(colorgenes, c) != NULL ? MAX_NUMBER_SAME_TYPE_COLOR : MAX_NUMBER_SAME_TYPE;
+                if (counter[lower] <= MAX_NUMBER) //get rid of modifiers that are too numerous - get rid of the first ones in the string (="oldest", the last ones looking from the end), because their influence on the parameter value is the smallest
                         simplified += c;
         }