Virtualized Grids and Multi-Column Layouts
You'll learn how to extend virtualization to two dimensions for table-like data and image galleries.
What Are Virtualized Grids?
While FixedSizeList handles vertical scrolling of single-column data, real applications often need grids—think data tables, photo galleries, or calendars. FixedSizeGrid from react-window virtualizes both rows and columns, rendering only the cells visible in the viewport. Instead of tracking just vertical scroll position, it monitors both axes simultaneously.
How FixedSizeGrid Works
Imagine a spreadsheet with 10,000 rows and 50 columns. Without virtualization, React would render 500,000 cells. FixedSizeGrid renders perhaps 20 rows × 10 columns (200 cells), dramatically reducing DOM nodes and render work.
You define:
columnCountandrowCount(total dimensions)columnWidthandrowHeight(fixed cell sizes)heightandwidth(viewport size)- A
Cellcomponent that receives{ columnIndex, rowIndex, style }
The grid calculates which cells to render based on scroll position, passing positioning via style.
Typical Use Cases
Data tables: Each cell displays a specific field from row data indexed by rowIndex and columnIndex.
Image galleries: Each cell renders a thumbnail from a two-dimensional array or a flattened list mapped to grid coordinates.
Calendars: Rows are weeks, columns are days—only visible weeks render.
The pattern remains familiar: provide total dimensions, cell sizes, and a render function. The library handles the complex scroll math.
Key Takeaway:
Use FixedSizeGrid to virtualize table-like data across both dimensions, rendering only visible cells for massive performance gains with large two-dimensional datasets.