python - How to plot a heatmap of a big matrix with matplotlib (45K * 446)

Question

I am trying to plot a heatmap of a big microarray dataset (45K rows per 446 columns). Using pcolor from matplotlib I am unable to do it because my pc goes easily out of memory (more than 8G)..

I'd prefer to use python/matplotlib instead of R for personal opinion.. Any way to plot heatmaps in an efficient way?

Thanks

Answer 1

我通过将矩阵下采样为更小的矩阵来解决。我决定尝试两种方法：

• 假设我想将 45k 行的矩阵下采样为 1k 行的矩阵，我每 45 行取一个行值
• 另一种方法是，将 45k 行下采样为 1k 行，将 45k 行分组为 1k 组（由 45 个相邻行组成），并将每组的平均值作为代表行

希望能帮助到你。