This PipeOp extracts time series features from functional columns.
For more details, see tsfeatures::tsfeatures(), which is called internally.
Parameters
The parameters are the parameters inherited from PipeOpTaskPreprocSimple,
as well as the following parameters:
features::character()
Function names which return numeric vectors of features. All features returned by these functions must be named if they return more than one feature. Default isc("frequency", "stl_features", "entropy", "acf_features").scale::logical(1)
IfTRUE, data is scaled to mean 0 and sd 1 before features are computed. Default isTRUE.trim::logical(1)
IfTRUE, data is trimmed bytrim_amountbefore features are computed. Values larger thantrim_amountin absolute value are set toNA. Default isFALSE.trim_amount::numeric(1)
Default level of trimming. Default is0.1.parallel::logical(1)
IfTRUE, the features are computed in parallel. Default isFALSE.multiprocess::any
The function from the future package to use for parallel processing. Default isfuture::multisession().na.action::any
A function to handle missing values. Default isstats::na.pass().
Naming
The new names generally append a _{feature} to the corresponding column name.
If a column was called "x" and the feature is "trend", the corresponding new column will
be called "x_trend".
Super classes
mlr3pipelines::PipeOp -> mlr3pipelines::PipeOpTaskPreproc -> mlr3pipelines::PipeOpTaskPreprocSimple -> PipeOpTSFeatures
Methods
Method new()
Initializes a new instance of this Class.
Usage
PipeOpTSFeatures$new(id = "fda.tsfeats", param_vals = list())Examples
task = tsk("fuel")
po_tsfeats = po("fda.tsfeats")
task_tsfeats = po_tsfeats$train(list(task))[[1L]]
task_tsfeats$data()
#> heatan h20 NIR_frequency NIR_nperiods NIR_seasonal_period NIR_trend
#> <num> <num> <num> <num> <num> <num>
#> 1: 26.7810 2.3000 1 0 1 0.9546922
#> 2: 27.4720 3.0000 1 0 1 0.9515743
#> 3: 23.8400 2.0002 1 0 1 0.4707549
#> 4: 18.1680 1.8500 1 0 1 0.4779991
#> 5: 17.5170 2.3898 1 0 1 0.9507122
#> ---
#> 125: 23.8340 2.1100 1 0 1 0.8778792
#> 126: 11.8050 1.6200 1 0 1 0.9854779
#> 127: 8.8315 1.4200 1 0 1 0.9887717
#> 128: 11.3450 1.4800 1 0 1 0.9562442
#> 129: 28.9940 2.5000 1 0 1 0.8044784
#> NIR_spike NIR_linearity NIR_curvature NIR_e_acf1 NIR_e_acf10
#> <num> <num> <num> <num> <num>
#> 1: 5.092984e-07 1.500377 -14.8389900 0.08526800 0.13922801
#> 2: 6.047324e-07 13.551188 -5.1228861 0.04333781 0.06594578
#> 3: 9.909298e-05 -8.992664 -3.7895780 -0.36900748 0.29693031
#> 4: 4.928023e-05 6.689307 -3.8887742 -0.10463336 0.19071080
#> 5: 6.500488e-07 14.045622 -4.3742672 0.18067626 0.09144276
#> ---
#> 125: 5.229296e-06 14.009846 -3.3695527 0.15252180 0.18735242
#> 126: 3.639621e-08 15.134587 -0.2588765 0.19727503 0.20779722
#> 127: 3.976177e-08 15.038312 -0.1175876 0.43423412 0.27618845
#> 128: 5.638686e-07 11.562683 -8.9000714 0.18090095 0.09866772
#> 129: 7.539249e-06 -12.302347 -4.8001860 -0.19585687 0.08648780
#> NIR_entropy NIR_x_acf1 NIR_x_acf10 NIR_diff1_acf1 NIR_diff1_acf10
#> <num> <num> <num> <num> <num>
#> 1: 0.4354773 0.9069888 6.3605955 -0.2832035 0.37528302
#> 2: 0.2939412 0.9448758 7.8253925 -0.4984864 0.36126344
#> 3: 0.6530189 0.2826531 2.0955036 -0.7039771 0.72744609
#> 4: 0.8348614 0.3648791 0.8041568 -0.5993595 0.72452181
#> 5: 0.3665353 0.9154091 6.6681478 -0.3014810 0.14010207
#> ---
#> 125: 0.3783612 0.8836760 5.9622264 -0.2487911 0.26465533
#> 126: 0.2542462 0.9616638 7.7381442 -0.5169663 0.65119007
#> 127: 0.0923419 0.9692264 7.7931485 -0.2257876 0.06959872
#> 128: 0.3504391 0.9495463 7.8098652 -0.3017364 0.17580374
#> 129: 0.2084854 0.7670506 5.7345844 -0.5114536 0.29116684
#> NIR_diff2_acf1 NIR_diff2_acf10 UVVIS_frequency UVVIS_nperiods
#> <num> <num> <num> <num>
#> 1: -0.5342972 0.6742914 1 0
#> 2: -0.6633171 0.5959168 1 0
#> 3: -0.7866512 0.9116270 1 0
#> 4: -0.7539921 1.1019418 1 0
#> 5: -0.6114086 0.4473425 1 0
#> ---
#> 125: -0.5366052 0.5611325 1 0
#> 126: -0.7609619 1.1970879 1 0
#> 127: -0.6044396 0.4192549 1 0
#> 128: -0.6131724 0.5282163 1 0
#> 129: -0.6604822 0.4903499 1 0
#> UVVIS_seasonal_period UVVIS_trend UVVIS_spike UVVIS_linearity
#> <num> <num> <num> <num>
#> 1: 1 0.9392390 1.503863e-06 10.552105
#> 2: 1 0.8320405 2.371308e-05 9.967401
#> 3: 1 0.7324108 1.782489e-05 8.067557
#> 4: 1 0.6591754 3.880087e-05 7.817542
#> 5: 1 0.7990575 2.365003e-05 9.251109
#> ---
#> 125: 1 0.4516040 9.644443e-05 6.383468
#> 126: 1 0.6914674 3.511260e-05 2.346512
#> 127: 1 0.8352613 8.696230e-06 1.220885
#> 128: 1 0.8974968 5.003478e-06 3.381766
#> 129: 1 0.7180091 6.170982e-05 9.689281
#> UVVIS_curvature UVVIS_e_acf1 UVVIS_e_acf10 UVVIS_entropy UVVIS_x_acf1
#> <num> <num> <num> <num> <num>
#> 1: 3.7969671 0.06234423 0.14352149 0.3430382 0.9268348
#> 2: 2.9073728 -0.12952783 0.08974614 0.4686135 0.8322983
#> 3: 1.6726579 -0.12604541 0.15486072 0.6220675 0.6844840
#> 4: 4.5641515 -0.21621644 0.12689752 0.6280994 0.5695714
#> 5: 3.9977271 -0.29746795 0.18589166 0.2662301 0.7274064
#> ---
#> 125: 3.7901989 -0.01400302 0.07658932 0.6899361 0.4432345
#> 126: 7.9806922 -0.21722795 0.16418124 0.6601140 0.5810357
#> 127: 9.8332792 -0.30735123 0.28247539 0.3755616 0.7721900
#> 128: 9.2982058 -0.18280897 0.07571742 0.4438987 0.8695996
#> 129: 0.3354961 -0.19728333 0.18618097 0.4433417 0.6908904
#> UVVIS_x_acf10 UVVIS_diff1_acf1 UVVIS_diff1_acf10 UVVIS_diff2_acf1
#> <num> <num> <num> <num>
#> 1: 7.025376 -0.3361110 0.3484453 -0.4994596
#> 2: 5.930650 -0.4541526 0.3318020 -0.6050046
#> 3: 3.739534 -0.4730931 0.3110636 -0.6450409
#> 4: 3.620160 -0.6062612 0.5471917 -0.7394589
#> 5: 4.917010 -0.6055341 0.5175728 -0.6662655
#> ---
#> 125: 1.452620 -0.4243995 0.2306197 -0.5893079
#> 126: 3.120828 -0.5156116 0.5919048 -0.6730118
#> 127: 4.787938 -0.6121137 0.6976292 -0.6873167
#> 128: 6.199741 -0.5181048 0.3134932 -0.6955121
#> 129: 4.286336 -0.4140705 0.3492405 -0.5288576
#> UVVIS_diff2_acf10
#> <num>
#> 1: 0.5223420
#> 2: 0.6000084
#> 3: 0.5107427
#> 4: 0.8566489
#> 5: 0.6403891
#> ---
#> 125: 0.3937492
#> 126: 1.0841297
#> 127: 0.9018225
#> 128: 0.6023695
#> 129: 0.5288989