How to split the data into train, test, and validation

[sauravsingh-couture]

• PyTorch-Forecasting version: 1.0.0
• PyTorch version: 2.0.1
• Python version: 3.8
• Operating System: Linux

I have data of 17 months, so my time_idx ranges from 0-16. There are total 3664 unique ids in my data. So my data has total rows 62288. I just want to split my data into train, val, and test.

• Train data ranges from 0 to 14 time_idx
• Val data should have 15 time_idx
• Test data should have 16 time_idx

So how would I achieve it from inbuilt class TimeSeriesDataset and inbuilt class method .from_dataset?

[ivanightingale]

You can refer to this example. See Cell 5 where trainingand validation are defined.