ROCKET
- class hana_ml.algorithms.pal.tsa.rocket.ROCKET(method=None, num_features=None, data_dim=None, random_seed=None)
RandOm Convolutional KErnel Transform (ROCKET) is an exceptionally efficient algorithm for time series classification. Unlike other proposed time series classification algorithms which attain excellent accuracy, ROCKET maintains its performance with a fraction of the computational expense by transforming input time series using random convolutional kernels.
- Parameters:
- methodstr, optional
The options are "MiniRocket" and "MultiRocket".
Defaults to "MiniRocket".
- num_featuresint, optional
Number of transformed features for each time series.
Defaults to 9996 when
method= "MiniRocket", 49728 whenmethod= "MultiRocket".- data_dimint, optional
Dimensionality of the multivariate time series.
1 means univariate time series and others for multivariate. Cannot be smaller than 1.
Defaults to 1.
- random_seedint, optional
0 indicates using machine time as seed.
Defaults to 0.
Examples
Example 1: Univariate time series fitted and transformed by MiniRocket Input DataFrame df:
>>> df.collect() RECORD_ID VAL_1 VAL_2 VAL_3 VAL_4 VAL_5 VAL_6 ... VAL_10 VAL_11 VAL_12 VAL_13 VAL_14 VAL_15 VAL_16 0 0 1.598 1.599 1.571 1.550 1.507 1.434 ... 1.117 1.024 0.926 0.828 0.739 0.643 0.556 1 1 1.701 1.671 1.619 1.547 1.475 1.391 ... 1.070 0.985 0.899 0.816 0.733 0.658 0.581 ... 10 10 1.614 1.574 1.557 1.521 1.460 1.406 ... 1.045 0.957 0.862 0.771 0.681 0.587 0.497 11 11 1.652 1.665 1.656 1.623 1.571 1.499 ... 1.155 1.058 0.973 0.877 0.797 0.704 0.609
Create an instance of ROCKET:
>>> ro = ROCKET(method="MiniRocket", random_seed=1)
Perform fit():
>>> ro.fit(data=df)
Model:
>>> ro.model_.collect() ID MODEL_CONTENT 0 -1 MiniRocket 1 0 {"SERIES_LENGTH":16,"NUM_CHANNELS":1,"BIAS_SIZ... 2 1 843045766098464,1.0396523357230486,2.005001093... ......
Make a transformation:
>>> result = ro.transform(data=df) >>> result.collect() ID STRING_CONTENT 0 0 {"NUM_FEATURES_PER_DATA":9996,"FEATURES":[{"DA... 1 1 ,0.375,0.875,0.125,0.5,1.0,0.25,0.75,1.0,0.375... ... 126 126 .0,0.0,0.75,0.0,0.375,1.0,0.0,0.75,1.0,0.125,0... 127 127 25,0.625,0.1875,0.375,0.75,0.0,0.625,0.0,0.0,0...
Example 2: Multivariate time series (with dimensionality 8) fitted and transformed by MultiRocket Input DataFrame df:
>>> df.collect() RECORD_ID VAL_1 VAL_2 VAL_3 VAL_4 VAL_5 VAL_6 ... VAL_10 VAL_11 VAL_12 VAL_13 VAL_14 VAL_15 VAL_16 0 0 1.645 1.646 1.621 1.585 1.540 1.470 ... 1.161 1.070 0.980 0.893 0.798 0.705 0.620 1 1 1.704 1.705 1.706 1.680 1.632 1.560 ... 1.186 1.090 0.994 0.895 0.799 0.702 0.605 ... 30 30 1.688 1.648 1.570 1.490 1.408 1.327 ... 1.011 0.930 0.849 0.768 0.687 0.606 0.524 31 31 1.708 1.663 1.595 1.504 1.411 1.318 ... 0.951 0.861 0.794 0.704 0.614 0.529 0.446
Create an instance of ROCKET:
>>> ro = ROCKET(method="multirocket", data_dim=8)
Perform fit():
>>> ro.fit(data=df)
Model:
>>> ro.model_.collect() ID MODEL_CONTENT 0 -1 MultiRocket 1 0 {"SERIES_LENGTH":16,"NUM_CHANNELS":8,"BIAS_SIZ... 2 1 HANNELS":[6]},{"ID":77,"CHANNELS":[1,4,7,6,5]}... ......
Make a transformation:
>>> result = ro.transform(data=df) >>> result.collect() ID STRING_CONTENT 0 0 {"NUM_FEATURES_PER_DATA":49728,"FEATURES":[{"D... 1 1 .5625,0.8125,0.125,0.6875,0.9375,0.5,0.6875,0.... .. ... ... 241 241 .857142,7.357142,12.333333,9.0,12.5,11.8,7.692... 242 242 1.0,3.0,3.0,5.0,3.0,-1.0,3.0,3.0,-1.0,3.0,-1.0...
- Attributes:
- model_DataFrame
Model content.
Methods
fit(data[, key])Fit the model to the training dataset.
Get the model metrics.
Get the score metrics.
transform(data[, key, thread_ratio])Transform time series based on a given ROCKET model fitted by fit().
- fit(data, key=None)
Fit the model to the training dataset.
- Parameters:
- dataDataFrame
Input data.
For univariate time series, each row represents one time series, while for multivariate time series, a fixed number of consecutive rows forms one time series, and that number is designated by the parameter
data_dimwhen initialize a ROCKET instance.- keystr, optional
The ID column.
Defaults to the first column of data if the index column of data is not provided. Otherwise, defaults to the index column of data.
- Returns:
- A fitted object of class "ROCKET".
- transform(data, key=None, thread_ratio=None)
Transform time series based on a given ROCKET model fitted by fit(). Hence, The data should be in the exact same format as that in fit(), especially the length and dimensionality of time series. The model_ used in transform comes from fit() as well.
- Parameters:
- dataDataFrame
Input data. For univariate time series, each row represents one time series, while for multivariate time series, a fixed number of consecutive rows forms one time series, and that number is designated by the parameter
data_dimwhen initialize a ROCKET instance.- keystr, optional
The ID column.
Defaults to the first column of data if the index column of data is not provided. Otherwise, defaults to the index column of data.
- thread_ratiofloat, optional
Adjusts the percentage of available threads to use, from 0 to 1. A value of 0 indicates the use of a single thread, while 1 implies the use of all possible current threads. Values outside the range will be ignored and this function heuristically determines the number of threads to use.
Defaults to 1.0.
- Returns:
- DataFrame
Features, structured as follows:
ID, type INTEGER, ROW_INDEX, indicates the ID of current row.
STRING_CONTENT, type NVARCHAR, transformed features in JSON format.
- get_model_metrics()
Get the model metrics.
- Returns:
- DataFrame
The model metrics.
- get_score_metrics()
Get the score metrics.
- Returns:
- DataFrame
The score metrics.
Inherited Methods from PALBase
Besides those methods mentioned above, the ROCKET class also inherits methods from PALBase class, please refer to PAL Base for more details.