breaking: raise error for gaps in series

[jmoralez]

Raises an error when the series' timestamps have gaps in them, are duplicated or don't match the specified frequency (freq argument).

[review-notebook-app]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[github-actions]

Experiment Results

Experiment 1: air-passengers

Description:

variable	experiment
h	12
season_length	12
freq	MS
level	None
n_windows	1

Results:

metric	timegpt-1	timegpt-1-long-horizon	SeasonalNaive	Naive
mae	12.6793	11.0623	47.8333	76
mape	0.027	0.0232	0.0999	0.1425
mse	213.936	199.132	2571.33	10604.2
total_time	9.1022	2.2177	0.0059	0.0047

Plot:

Experiment 2: air-passengers

Description:

variable	experiment
h	24
season_length	12
freq	MS
level	None
n_windows	1

Results:

metric	timegpt-1	timegpt-1-long-horizon	SeasonalNaive	Naive
mae	58.1031	58.4587	71.25	115.25
mape	0.1257	0.1267	0.1552	0.2358
mse	4040.21	4110.79	5928.17	18859.2
total_time	0.6477	1.2845	0.0045	0.0042

Plot:

Experiment 3: electricity-multiple-series

Description:

variable	experiment
h	24
season_length	24
freq	H
level	None
n_windows	1

Results:

metric	timegpt-1	timegpt-1-long-horizon	SeasonalNaive	Naive
mae	178.293	268.13	269.23	1331.02
mape	0.0234	0.0311	0.0304	0.1692
mse	121589	219485	213677	4.68961e+06
total_time	0.4795	2.1672	0.0055	0.0053

Plot:

Experiment 4: electricity-multiple-series

Description:

variable	experiment
h	168
season_length	24
freq	H
level	None
n_windows	1

Results:

metric	timegpt-1	timegpt-1-long-horizon	SeasonalNaive	Naive
mae	465.497	346.984	398.956	1119.26
mape	0.062	0.0437	0.0512	0.1583
mse	835021	403787	656723	3.17316e+06
total_time	0.6089	1.3617	0.006	0.0055

Plot:

Experiment 5: electricity-multiple-series

Description:

variable	experiment
h	336
season_length	24
freq	H
level	None
n_windows	1

Results:

metric	timegpt-1	timegpt-1-long-horizon	SeasonalNaive	Naive
mae	558.673	459.757	602.926	1340.95
mape	0.0697	0.0565	0.0787	0.17
mse	1.22723e+06	739114	1.61572e+06	6.04619e+06
total_time	2.3025	2.4547	0.0058	0.0054

Plot:

[elephaint]

Two comments, for the comment about the addition to the notebooks in the capabilities, that one might also be appropriate in the tutorial 12_irregular

[elephaint]

Maybe we should add a callout to this capabilities notebook at the bottom stating that TimeGPT doesn't allow gaps in the timestamps? E.g.

"Make sure there are no gaps in your time series data. This means that even if the chosen frequency is irregular, you should still make sure you provide a value for every irregular timestamp in the data. For example, if your frequency is "B" (business day), there can't be a gap (missing datapoint) between two consecutive business days."

Edit: perhaps add a similar comment also to the beginning or end of the tutorial notebook 12_irregular.

[jmoralez]

We already have that in the data requirements notebook

When using TimeGPT, the data cannot contain missing values. This means that for every series, there should be no gaps in the timestamps and no missing values in the target variable.

[elephaint]

Sure, but repetition is the key to education? 😆

[jmoralez]

Yeah that may help, but I'd prefer to add a link to that section instead, otherwise we'll have to remember to change that in every place we set it and will most likely miss some.

[elephaint]

Small non-blocking comments