如何优化二维数组特定值计数的双层循环？以棋盘空白格统计为例

Great question! Let's break down how we can simplify and optimize your blank space counting task—while keeping things efficient.

First, a quick reality check: since we have to check every cell to count blank spaces, we can't escape an O(width*height) time complexity. But we can make the code cleaner and faster by leveraging Python's built-in tools, which are implemented in optimized C under the hood.

Here are a few better approaches:

1. Use sum() with row-wise count() #

This is the most straightforward and readable method. We count the number of 0s in each row, then sum those counts:

blank_spaces = sum(row.count(0) for row in game.__board_state__)

The row.count(0) method is optimized, and using a generator expression with sum() avoids creating unnecessary intermediate lists, keeping memory usage low.

2. Flatten the board with itertools.chain #

If you prefer working with a flattened version of the board, itertools.chain can turn your 2D list into a single iterable of cells. Then you can either count directly or use a memory-efficient generator with sum():

from itertools import chain

# Option 1: Flatten to a list and count (uses more memory)
blank_spaces = list(chain.from_iterable(game.__board_state__)).count(0)

# Option 2: Generator expression (memory-light)
blank_spaces = sum(1 for cell in chain.from_iterable(game.__board_state__) if cell == 0)

3. Pre-maintain a counter (for frequent queries) #

If you need to count blank spaces very often (like in every game loop iteration), the ultimate optimization is to track the count dynamically instead of recalculating it every time:

• When initializing the game:

  self.blank_spaces = self.width * self.height
  

• When placing a non-zero piece on the board:

  self.blank_spaces -= 1
  

• When resetting a cell back to 0:

  self.blank_spaces += 1
  

Now you can get the blank space count in O(1) time just by accessing game.blank_spaces—no loops needed at all!

All these approaches are better than hand-writing nested for loops, as they're either more readable, faster, or both.

内容的提问来源于stack exchange，提问作者Suhail Gupta