Multi-dimensional wrapper to compute the non-normalized (i.e., without z-normalization multi-dimensional matrix profile and multi-dimensional matrix profile index for a given window within the times series or sequence that is denoted by the `start` index. Essentially, this is a conv
(
start,
T_A,
T_B,
m,
excl_zone,
T_B_subseq_isfinite,
p=2.0,
include=None,
discords=False,
)
| 458 | |
| 459 | |
| 460 | def _get_first_maamp_profile( |
| 461 | start, |
| 462 | T_A, |
| 463 | T_B, |
| 464 | m, |
| 465 | excl_zone, |
| 466 | T_B_subseq_isfinite, |
| 467 | p=2.0, |
| 468 | include=None, |
| 469 | discords=False, |
| 470 | ): |
| 471 | """ |
| 472 | Multi-dimensional wrapper to compute the non-normalized (i.e., without |
| 473 | z-normalization multi-dimensional matrix profile and multi-dimensional matrix |
| 474 | profile index for a given window within the times series or sequence that is denoted |
| 475 | by the `start` index. Essentially, this is a convenience wrapper around |
| 476 | `_multi_mass_absolute`. This is a convenience wrapper for the |
| 477 | `_maamp_multi_distance_profile` function but does not return the multi-dimensional |
| 478 | matrix profile subspace. |
| 479 | |
| 480 | Parameters |
| 481 | ---------- |
| 482 | start : int |
| 483 | The window index to calculate the first multi-dimensional matrix profile, |
| 484 | multi-dimensional matrix profile indices, and multi-dimensional subspace. |
| 485 | |
| 486 | T_A : numpy.ndarray |
| 487 | The time series or sequence for which the multi-dimensional matrix profile, |
| 488 | multi-dimensional matrix profile indices, and multi-dimensional subspace will be |
| 489 | returned |
| 490 | |
| 491 | T_B : numpy.ndarray |
| 492 | The time series or sequence that contains your query subsequences |
| 493 | |
| 494 | m : int |
| 495 | Window size |
| 496 | |
| 497 | excl_zone : int |
| 498 | The half width for the exclusion zone relative to the `start`. |
| 499 | |
| 500 | T_B_subseq_isfinite : numpy.ndarray |
| 501 | A boolean array that indicates whether a subsequence in `T_B` contains a |
| 502 | `np.nan`/`np.inf` value (False) |
| 503 | |
| 504 | p : float, default 2.0 |
| 505 | The p-norm to apply for computing the Minkowski distance. Minkowski distance is |
| 506 | typically used with `p` being 1 or 2, which correspond to the Manhattan distance |
| 507 | and the Euclidean distance, respectively. |
| 508 | |
| 509 | include : numpy.ndarray, default None |
| 510 | A list of (zero-based) indices corresponding to the dimensions in `T` that |
| 511 | must be included in the constrained multidimensional motif search. |
| 512 | For more information, see Section IV D in: |
| 513 | |
| 514 | `DOI: 10.1109/ICDM.2017.66 \ |
| 515 | <https://www.cs.ucr.edu/~eamonn/Motif_Discovery_ICDM.pdf>`__ |
| 516 | |
| 517 | discords : bool, default False |