TS features selector

This module belongs to the "Time Series processing" category of modules.

Description

This module takes a Pandas dataframe as an input and outputs 2 arrays : the first is the features list and the second is the labels list. This is used for time series datasets. Just add the name of the column you want to consider as features and labels to the corresponding parameters .

The ports respectively output : a list of features and a list of labels.

Parameters

As parameters , we can define :

• the window size

• the features dimension

• the number of classes.

Columns' names are automatically retrieved from the datasets inside the drag and drop area. They will appear within the combo-box and are able to be selected in batch.

However , we can also choose to specify them one by one through manual setting.

Furthermore, we also have the options to encode the labels or not with a One Hot Encoder.

This interesting article provides links of some repositories for downloading datasets including time series datasets (for free).