HyperFormula (v3.4.0)

View as Markdown

# Performance

We implemented various techniques to boost the performance of HyperFormula. In some cases, turning them on or off might increase the performance of your app. Below we provide a number of tips on how to speed it up.

# Loading multiple sheets

Build the engine once with all the data instead of adding sheets one by one. HyperFormula's buildFromSheets method takes every sheet in a single call and resolves the whole dependency graph once:

const hf = HyperFormula.buildFromSheets({
  Sheet1: [ ['1', '=Sheet2!A1'] ],
  Sheet2: [ ['10'] ],
}, { licenseKey: 'gpl-v3' })

Loading the same data incrementally, with an addSheet and a setSheetContent call per sheet, is much slower, and the gap widens with every sheet added. Each setSheetContent call recalculates every loaded cell that depends on the sheet it just filled, so when the sheets already loaded reference the one being added, the work grows with each step. In a test with 500 cross-referencing sheets, the per-sheet loop was more than a hundred times slower than a single buildFromSheets call.

addSheet adds to this whenever the formulas already loaded point at the sheet being added: it marks all the cells and ranges of that sheet as dirty and recalculates on its own; in the test above that roughly doubled the cost of each step. Registering all the sheet names upfront removes that share, because there is nothing to recalculate yet when the names are registered, but it does not remove the growth.

When the data is not known upfront and sheets have to be added at runtime, group the operations into a batch so that the recalculation runs once for the whole group instead of once per operation, and keep whatever renders your data from reading the engine until the batch ends: the methods that read cell values throw while the evaluation is suspended. See suspending automatic recalculations below.

# Order of loading

Since version 3.1.1, a formula that references a sheet which has not been added yet evaluates to #REF! but keeps a live reference: adding that sheet later repairs the formula, with no re-parsing and no rebuild on your side. Load the sheets in whatever order is convenient.

const hf = HyperFormula.buildFromSheets({
  Hub: [ ['=Later!A1+1'] ],  // #REF! for now
}, { licenseKey: 'gpl-v3' })

hf.addSheet('Later')                                  // Hub!A1 is 1
hf.setSheetContent(hf.getSheetId('Later'), [ ['41'] ])  // Hub!A1 is 42

This also means that ordering the inserts by dependency is not a fix for the cost described above. It helps only as long as every reference happens to point the same way, and a single reference pointing back puts you on the slow path again with nothing to signal it.

# Passing the engine to other libraries

A library you hand HyperFormula to may accept either the HyperFormula class or a ready instance. Given the class, it builds the engine on its own terms, and an integration that receives your sheets one at a time will add them the same way – the slow path above. Building the instance yourself with buildFromSheets and passing that instead takes the decision out of its hands.

# VLOOKUP/MATCH

If you are planning to use VLOOKUP or MATCH heavily in your app, you may consider enabling the useColumnIndex flag in the HyperFormula configuration. It will increase memory usage but can significantly improve the performance of these two functions, especially when running on unsorted or very large data sets. The column index will not be used despite the option useColumnIndex enabled when using wildcards or regular expressions.

Leaving this option disabled will cause the engine to use binary search when dealing with sorted data, and the naive approach otherwise.

# Address mapping strategies

HyperFormula uses two approaches to store the mapping of cell addresses in order to optimize memory usage. The choice of the strategy is made independently for each sheet. The chooseAddressMappingPolicy option allows for changing the way the strategy will be chosen.

You may use one of three built-in policies:

  • AlwaysDense – uses dense mapping for each sheet. This policy is particularly useful when the spreadsheet is a densely filled rectangle.
  • AlwaysSparse – uses sparse mapping for each sheet. This approach is useful when in your spreadsheet/dataset there are relatively few cells filled, but located very far from each other.
  • DenseSparseChooseBasedOnThreshold – the choice is made based on the fill ratio of the sheet. Let the engine choose the best strategy for you.

# Lazy transformation cleanup

Structural operations (adding/removing rows/columns, moving cells) create transformations that are applied lazily to formulas. Over time, these transformations accumulate in memory. HyperFormula automatically flushes them when their count reaches the maxPendingLazyTransformations threshold (default: 50).

You can tune this setting to balance memory usage and CPU overhead:

  • Lower values (e.g., 10) reduce peak memory usage but trigger cleanup more frequently, adding slight CPU overhead per flush.
  • Higher values (e.g., 200) reduce the frequency of cleanup but allow more memory to accumulate between flushes.
  • The default of 50 works well for most use cases.
const hf = HyperFormula.buildEmpty({
  licenseKey: 'gpl-v3',
  maxPendingLazyTransformations: 100,
})

# Suspending automatic recalculations

By default, HyperFormula recalculates formulas after every change. However, due to the fact that we store the graph of dependencies between cells in the sheet, we recalculate only the cells affected by the update.

Sometimes, a simple change can cause recalculation of a large part of the sheet, e.g., when the modified cell is at the very beginning of the dependency chain or when there are many volatile functions in the worksheet. In such a case you may want to postpone the recalculation.

The first option is to call suspendEvaluation before making changes and resumeEvaluation at a convenient moment.

The second option is to pass the callback function with multiple operations to a batch function. Recalculation will be suspended before performing operations and resumed after them. In cases where you perform operations which may not cause a recalculation but only change the shape of the worksheet, like addRows, removeRows, or moveColumns , we do not recommend suspending recalculation, as this may have a slightly negative impact on performance.

← Known limitations →