Historical Forecast

1. TimeGPT Historical Forecast

When generating a forecast, sometimes you might be interested in forecasting the historical observations. These predictions, known as fitted values, can help you better understand and evaluate a model’s performance over time.

TimeGPT has a method for generating fitted values, and users can call it from nixtlar. This vignette will explain how to do this. It assumes you have already set up your API key. If you haven’t done this, please read the Get Started vignette first.

2. Load data

For this vignette, we’ll use the electricity consumption dataset that is included in nixtlar, which contains the hourly prices of five different electricity markets.

df <- nixtlar::electricity
head(df)
#>   unique_id                  ds     y
#> 1        BE 2016-10-22 00:00:00 70.00
#> 2        BE 2016-10-22 01:00:00 37.10
#> 3        BE 2016-10-22 02:00:00 37.10
#> 4        BE 2016-10-22 03:00:00 44.75
#> 5        BE 2016-10-22 04:00:00 37.10
#> 6        BE 2016-10-22 05:00:00 35.61

3. Forecast historical data

To generate a forecast for the historical data, use nixtlar::nixtla_client_historic, which should include the following parameters:

• df: The time series data, provided as a data frame, tibble, or tsibble. It must include at least two columns: one for the timestamps and one for the observations. The default names for these columns are ds and y. If your column names are different, specify them with time_col and target_col, respectively. If you are working with multiple series, you must also include a column with unique identifiers. The default name for this column is unique_id; if different, specify it with id_col.
• level: The prediction intervals for the forecast.

nixtla_client_fitted_values <- nixtla_client_historic(df, level = c(80,95))
#> Frequency chosen: h
head(nixtla_client_fitted_values)
#>   unique_id                  ds  TimeGPT TimeGPT-lo-80 TimeGPT-lo-95
#> 1        BE 2016-10-24 00:00:00 46.99983      44.54957      44.09014
#> 2        BE 2016-10-24 01:00:00 42.07846      35.73659      34.54749
#> 3        BE 2016-10-24 02:00:00 40.89080      31.17443      29.35261
#> 4        BE 2016-10-24 03:00:00 38.80745      22.75283      19.74259
#> 5        BE 2016-10-24 04:00:00 38.59341      18.62503      14.88096
#> 6        BE 2016-10-24 05:00:00 36.58477      16.03442      12.18123
#>   TimeGPT-hi-80 TimeGPT-hi-95
#> 1      49.45010      49.90952
#> 2      48.42033      49.60943
#> 3      50.60717      52.42899
#> 4      54.86207      57.87231
#> 5      58.56179      62.30587
#> 6      57.13511      60.98830

Notice that there are no fitted values for some of the initial observations. This is because TimeGPT requires a minimum number of values to generate a forecast for the historical data.

All the fitted values are generated using a rolling window, meaning that the fitted value for observation \(T\) is generated using the first \(T-1\) observations.