So we see from the density plot that the most dense part happens to be intersection of the x, y histograms where the highest bar meet. But suppose we only have the x, y histograms, I think it's not sufficient condition to infer the densest part to the intersection of the highest points, because the histograms counts the total on the row and column, the distribution across the row or column, however, is not clear.
I want to ask if it's possible to construct a density plot, if we can count the total from any line across a plot, not just the row or column, but taken from any angle. It would kinda resemble a sudoku puzzle in a way? What area of math would this fall under? Linear algebra (solving for unknown vars with enough conditions)?
Background:
I've come up with this hypothetical question due to echarts seeming to lack a 2d density plot feature (I probably need to use the Heatmap). The below plot is taken from Plotly (though I can't use it because it lacks other features).
