NSTableView: Finding out column widths using a Monte Carlo Simulation?

Question

NSTableView declares this optional delegate method:

- (CGFloat)tableView:(NSTableView *)tableView sizeToFitWidthOfColumn:(NSInteger)column;

Implementing it is easy, but when your table is filled with thousands of entries, you start to see performance issues, and calling this method makes the app unresponsive.

Here is what the documentation states regarding implementing the method:

Implement this method if you want to control how wide a column is made when the user double clicks on the resize divider. By default, NSTableView iterates every row in the table, accesses a cell via preparedCellAtRow:column:, and requests the "cellSize" to find the appropriate largest width to use.

For large row counts, a monte carlo simulation is done instead of interating every row. For performance and accurate results, it is recommended that this method is implemented when using large tables.

So i began crawling the web, searching for "Monte Carlo Simulations".

I now have a brief understanding of what a Monte-Carlo simulation is, but i don't understand how it can apply to this particular problem.

I mean, the max content size is something totally random. It won't follow any equation or rule, it's solely based on the content that will be inserted into the cells, so how does a Monte-Carlo method apply here, how will it help me find the adequate content size ?

Answer 1

Monte Carlo: Instead of measuring each cell of the column, only a random set of cells are measured and the maximum measurement found is considered the maximum width of the column. Of course this will not result in the maximum cell width. But it might be pretty close most of the time. And given the purpose of the method, and that is to resize a column on double clicking the resize divider, the error is no big deal. If you want the precise value, you shouldn't count on this method.

Why would you want to implement such method btw?