infer_unknown_changepoints {BCT} | R Documentation |

## Inferring the number of change-points and their locations.

### Description

This function implements the Metropolis-Hastings sampling algorithm for inferring the number of change-points and their locations.

### Usage

```
infer_unknown_changepoints(
input_data,
l_max,
depth,
alphabet,
iters,
fileName = NULL
)
```

### Arguments

`input_data` |
the sequence to be analysed. |

`l_max` |
maximum number of change-points. |

`depth` |
maximum memory length. |

`alphabet` |
symbols appearing in the sequence. |

`iters` |
number of iterations; for more information see Lungu et al. (2022). |

`fileName` |
file path for storing the results. |

### Value

return a list object which includes:

`number_changes` |
sampled number of change-points. |

`positions` |
sampled locations of the change-points. |

`acceptance_prob` |
the empirical acceptance ratio. |

### See Also

### Examples

```
# Use as an example the three_changes dataset.
# Run the function with 5 change-points, a maximum depth of 5 and the [0,1,2] alphabet.
# The sampler is run for 100 iterations
output <- infer_unknown_changepoints(three_changes, 5, 5, c("012"), 100, fileName = NULL)
# If the fileName is not set to NULL,
# the output file will contain on each line the sampled number of change-points
# and the associated sampled locations of the change-points.
```

[Package

*BCT*version 1.2 Index]