Better greed

agents/greed-v1.rhai

@@ -0,0 +1,161 @@
+// Return a random valid action on the given board.
+// Used as a last resort.
+fn random_action(board) {
+	let symb = rand_symb();
+	let pos = rand_int(0, 10);
+	let action = Action(symb, pos);
+
+	while !board.can_play(action) {
+		let symb = rand_symb();
+		let pos = rand_int(0, 10);
+		action = Action(symb, pos);
+	}
+
+	return action
+}
+
+
+/// Returns an array of (idx, f32) for each empty slot in the board.
+/// - idx is the index of this slot
+/// - f32 is the "influence of" this slot
+fn compute_influence(board) {
+	// Fill all empty slots with fives and compute starting value
+	let filled = board;
+	for i in filled.free_spots_idx() {
+		filled[i] = 5;
+	}
+
+	// Compute the value of the filled board
+	let base = filled.evaluate();
+
+	// Exit early if the board is invalid.
+	// This is usually caused by zero-division.
+	if (base == ()) {
+		return [];
+	}
+
+	// Increase each slot's value by 1
+	// and record the effect on the expression's total value.
+	//
+	// `influence` is an array of (slot_idx, value)
+	let influence = [];
+	for i in 0..board.size() {
+		let slot = board[i];
+
+		// Ignore slots that are not empty
+		if slot != "" {
+			continue
+		}
+
+		// Don't assign directly to `filled`,
+		// we want to keep it full of fives.
+		// Assigning to `b` make a copy of the board.
+		let b = filled;
+		b[i] = 6;
+
+		influence.push([i, b.evaluate() - base]);
+	}
+
+
+	// Sort by increasing absolute score
+	influence.sort(|a, b| {
+		let a_abs = a[1].abs();
+		let b_abs = b[1].abs();
+
+		// Returns...
+		//  1 if positive (a_abs > b_abs),
+		// -1 if negative,
+		//  0 if equal
+		return sign(a_abs - b_abs);
+	});
+
+	return influence;
+}
+
+fn place_number(board, minimize) {
+	let numbers = [0,1,2,3,4,5,6,7,8,9];
+	let available_numbers = numbers.retain(|x| board.contains(x));
+
+	let influence = compute_influence(board);
+
+	// Stupid edge cases, fall back to random
+	if influence.len() == 0 || available_numbers.len() == 0 {
+		return random_action(board);
+	}
+
+	// Get the most influential position
+	let pos = influence[-1][0];
+	let val = influence[-1][1];
+
+	// Pick the number we should use,
+	// This is always either the largest
+	// or the smallest number available to us.
+	let symbol = 0;
+	if minimize {
+		if val > 0 {
+			symbol = available_numbers[0];
+		} else {
+			symbol = available_numbers[-1];
+		}
+	} else {
+		if val > 0 {
+			symbol = available_numbers[-1];
+		} else {
+			symbol = available_numbers[0];
+		}
+	}
+
+	return Action(symbol, pos);
+}
+
+fn place_op(board, minimize) {
+	let ops = ["+", "-", "*", "/"];
+	let available_ops = ops.retain(|x| board.contains(x));
+
+ 	// Place operations first,
+	// they matter much more than numbers
+	let give_up = 10;
+ 	if !available_ops.is_empty() {
+		let aa = available_ops.rand_shuffle();
+		let pos = rand_int(0, 10);
+		let action = Action(aa[0], pos);
+
+		while !board.can_play(action) {
+			let pos = rand_int(0, 10);
+			action = Action(aa[0], pos);
+
+			// In case there are no valid operator moves
+			give_up -= 1;
+			if give_up == 0 { break }
+		}
+
+		return action
+	}
+
+	// Could not place an operation
+	return ();
+}
+
+// Main step function (shared between min and max)
+fn greed_step(board, minimize) {
+	let action = place_op(board, minimize);
+
+	// We could not place an op, so place a number
+	if action == () {
+		action = place_number(board, minimize);
+	}
+
+	// Prevent invalid moves, random fallback
+	if board.can_play(action) { return action; }
+	return random_action(board);
+}
+
+// Minimizer step
+fn step_min(board) {
+	greed_step(board, true)
+}
+
+// Maximizer step
+fn step_max(board) {
+	greed_step(board, false)
+}