This function "unconfounds" the candidate matrix. At each time point and for each location, we have the number of difference series which resulted in a changepoint. The location with the largest count is assumed to be the location where the changepoint occurs. Assignment of changepoints should then proceed iteratively, where each new changepoint is assigned based on the current highest count.

1 | ```
unconfoundCandidateMatrix(candidate, pairs, statistics, data, period, avgDiff)
``` |

`candidate` |
The candidate matrix, as computed by ?createCandidateMatrix. |

`pairs` |
The list object whose ith element specifies the neighboring locations to the ith location. |

`statistics` |
The time x (number of pairs) matrix of SNHT statistics computed for each difference series. |

`data` |
The data.frame containing the observations, restructured as in pairwiseSNHT. So, the first column should be time, and the other columns should be named with the locations and contain the observed values at each location. |

`period` |
The SNHT works by calculating the mean of the data on the previous period observations and the following period observations. Thus, this argument controls the window size for the test statistics. |

`avgDiff` |
A matrix containing the average differences between time series pairs. Generally this is created within pairwiseSNHT(). |

A list of two elements. The first element contains the data after the breaks have been removed. The second element is a data.frame with information regarding the detected changepoints (or NULL if none are found).

Questions? Problems? Suggestions? Tweet to @rdrrHQ or email at ian@mutexlabs.com.

Please suggest features or report bugs with the GitHub issue tracker.

All documentation is copyright its authors; we didn't write any of that.