---
title: minimumSequenceLength
framework: createmlcomponents
role: symbol
role_heading: Instance Property
path: createmlcomponents/timeseriesclassifierconfiguration/minimumsequencelength
---

# minimumSequenceLength

The minimum number of samples required to produce a classification.

## Declaration

```swift
var minimumSequenceLength: Int
```

## Discussion

Discussion This configuration parameter controls the size of the model. Set it to the largest value that is reasonable for your task. For example if your input is accelerometer data sampled at 100Hz and you want to recognize actions, the minimum action duration may be one second, so the minimum sequence length is 100Hz * 1s = 100. Defaults to 100.

## See Also

### Inspecting a time series classifier configuration

- [batchSize](createmlcomponents/timeseriesclassifierconfiguration/batchsize.md)
- [earlyStoppingIterationCount](createmlcomponents/timeseriesclassifierconfiguration/earlystoppingiterationcount.md)
- [earlyStoppingTolerance](createmlcomponents/timeseriesclassifierconfiguration/earlystoppingtolerance.md)
- [learningRate](createmlcomponents/timeseriesclassifierconfiguration/learningrate.md)
- [maximumIterationCount](createmlcomponents/timeseriesclassifierconfiguration/maximumiterationcount.md)
- [maximumSequenceLength](createmlcomponents/timeseriesclassifierconfiguration/maximumsequencelength.md)
- [randomSeed](createmlcomponents/timeseriesclassifierconfiguration/randomseed.md)