Electricity demand use case #321

marcopeix · 2024-04-29T20:00:22Z

Compared TimeGPT and N-HiTS for a problem of electricity demand forecasting using this dataset from the Monash Repository, based on this tutorial from SkForecast.

TimeGPT is more accurate and faster than N-HiTS.

Include use case for intermittent demand. TimeGPT performs 14% better than the top statistical model for intermittent forecasting.

review-notebook-app · 2024-04-29T20:00:26Z

Check out this pull request on

See visual diffs & provide feedback on Jupyter Notebooks.

Powered by ReviewNB

github-actions · 2024-04-29T20:03:03Z

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	5.4198	8.8487	0.0084	0.0048

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	11.0278	12.6738	0.0056	0.0047

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	142.394	196.363	269.23	1331.02
mape	0.0203	0.0234	0.0304	0.1692
mse	63464.7	123119	213677	4.68961e+06
total_time	12.306	12.9878	0.0076	0.0068

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	522.427	353.528	398.956	1119.26
mape	0.069	0.0454	0.0512	0.1583
mse	966294	422332	656723	3.17316e+06
total_time	14.7884	12.6789	0.0074	0.0068

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	478.362	361.033	602.926	1340.95
mape	0.0622	0.046	0.0787	0.17
mse	805039	441118	1.61572e+06	6.04619e+06
total_time	13.4245	12.679	0.0075	0.0071

Plot:

github-actions · 2024-04-29T20:38:56Z

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	19.0203	19.6633	0.0083	0.0046

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	23.5662	26.1395	0.0057	0.0047

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	142.394	196.363	269.23	1331.02
mape	0.0203	0.0234	0.0304	0.1692
mse	63464.7	123119	213677	4.68961e+06
total_time	20.9469	20.1594	0.0077	0.0067

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	522.427	353.528	398.956	1119.26
mape	0.069	0.0454	0.0512	0.1583
mse	966294	422332	656723	3.17316e+06
total_time	27.8454	19.63	0.0074	0.0069

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	478.362	361.033	602.926	1340.95
mape	0.0622	0.046	0.0787	0.17
mse	805039	441118	1.61572e+06	6.04619e+06
total_time	28.5248	19.1786	0.0076	0.0071

Plot:

mergenthaler · 2024-04-29T20:50:04Z

Not sure how to handle the dataset, since I had to download it locally. Should I add it to the transfer-learning repo and use the GitHub raw link after?
Yes, please.

github-actions · 2024-04-30T14:15:43Z

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	3.0409	2.7428	0.0085	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	5.1086	6.1719	0.0058	0.0048

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	142.394	196.363	269.23	1331.02
mape	0.0203	0.0234	0.0304	0.1692
mse	63464.7	123119	213677	4.68961e+06
total_time	6.9191	9.9466	0.0081	0.007

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	522.427	353.528	398.956	1119.26
mape	0.069	0.0454	0.0512	0.1583
mse	966294	422332	656723	3.17316e+06
total_time	9.8425	9.0534	0.0076	0.0071

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	478.362	361.033	602.926	1340.95
mape	0.0622	0.046	0.0787	0.17
mse	805039	441118	1.61572e+06	6.04619e+06
total_time	11.5873	11.4957	0.0075	0.007

Plot:

github-actions · 2024-04-30T14:23:52Z

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	2.3735	2.4545	0.0092	0.005

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	2.6259	2.5252	0.0063	0.005

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	142.394	196.363	269.23	1331.02
mape	0.0203	0.0234	0.0304	0.1692
mse	63464.7	123119	213677	4.68961e+06
total_time	2.5802	2.7442	0.0082	0.007

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	522.427	353.528	398.956	1119.26
mape	0.069	0.0454	0.0512	0.1583
mse	966294	422332	656723	3.17316e+06
total_time	3.5912	2.9026	0.0076	0.0072

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	478.362	361.033	602.926	1340.95
mape	0.0622	0.046	0.0787	0.17
mse	805039	441118	1.61572e+06	6.04619e+06
total_time	4.8539	3.0167	0.0078	0.0072

Plot:

mergenthaler · 2024-04-30T22:09:07Z

@marcopeix, is this ready to be reviewed by Azul and merged?

…mand_example

github-actions · 2024-05-01T17:30:04Z

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	213.935	199.132	2571.33	10604.2
total_time	3.1444	2.2909	0.0087	0.0049

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	2.7376	2.5569	0.006	0.0049

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	142.394	196.363	269.23	1331.02
mape	0.0203	0.0234	0.0304	0.1692
mse	63464.7	123119	213677	4.68961e+06
total_time	2.4477	4.8225	0.0079	0.007

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	522.427	353.528	398.956	1119.26
mape	0.069	0.0454	0.0512	0.1583
mse	966294	422332	656723	3.17316e+06
total_time	3.5015	3.8879	0.0076	0.0071

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	478.362	361.033	602.926	1340.95
mape	0.0622	0.046	0.0787	0.17
mse	805039	441118	1.61572e+06	6.04619e+06
total_time	5.7799	5.1059	0.0077	0.0074

Plot:

marcopeix · 2024-05-01T19:33:48Z

@AzulGarza , ready for review! 😄

github-actions · 2024-05-01T19:57:51Z

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	19.4155	19.8653	0.0084	0.0048

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	21.4148	14.6504	0.0057	0.0047

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	142.394	196.363	269.23	1331.02
mape	0.0203	0.0234	0.0304	0.1692
mse	63464.7	123119	213677	4.68961e+06
total_time	15.6412	18.7956	0.0079	0.0071

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	522.427	353.528	398.956	1119.26
mape	0.069	0.0454	0.0512	0.1583
mse	966294	422332	656723	3.17316e+06
total_time	21.8367	19.5205	0.0073	0.0067

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	478.362	361.033	602.926	1340.95
mape	0.0622	0.046	0.0787	0.17
mse	805039	441118	1.61572e+06	6.04619e+06
total_time	22.1962	21.533	0.0076	0.007

Plot:

github-actions · 2024-05-02T14:35:17Z

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	213.935	199.132	2571.33	10604.2
total_time	19.0402	12.5607	0.0089	0.0054

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	4040.22	4110.79	5928.17	18859.2
total_time	8.8411	11.6847	0.0058	0.0049

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	142.394	196.363	269.23	1331.02
mape	0.0203	0.0234	0.0304	0.1692
mse	63464.8	123119	213677	4.68961e+06
total_time	12.5136	13.8804	0.0081	0.007

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	522.427	353.528	398.956	1119.26
mape	0.069	0.0454	0.0512	0.1583
mse	966295	422332	656723	3.17316e+06
total_time	15.9889	16.4344	0.0076	0.007

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	478.362	361.033	602.926	1340.95
mape	0.0622	0.046	0.0787	0.17
mse	805038	441118	1.61572e+06	6.04619e+06
total_time	17.4498	13.7625	0.0078	0.0072

Plot:

nbs/docs/5_use_cases/3_electricity_demand.ipynb

AzulGarza

thanks @marcopeix! i left a couple of comments :)

github-actions · 2024-05-03T13:18:32Z

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	5.1378	3.644	0.0083	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	4.5861	3.2529	0.0057	0.0048

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	142.394	196.363	269.23	1331.02
mape	0.0203	0.0234	0.0304	0.1692
mse	63464.7	123119	213677	4.68961e+06
total_time	4.3862	7.6641	0.0079	0.007

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	522.427	353.528	398.956	1119.26
mape	0.069	0.0454	0.0512	0.1583
mse	966294	422332	656723	3.17316e+06
total_time	8.0733	19.4855	0.0073	0.0069

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	478.362	361.033	602.926	1340.95
mape	0.0622	0.046	0.0787	0.17
mse	805039	441118	1.61572e+06	6.04619e+06
total_time	9.2709	13.74	0.0076	0.007

Plot:

github-actions · 2024-05-03T14:36:52Z

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	4.2097	3.3494	0.009	0.0049

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	5.592	5.2048	0.0056	0.0046

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	142.394	196.363	269.23	1331.02
mape	0.0203	0.0234	0.0304	0.1692
mse	63464.7	123119	213677	4.68961e+06
total_time	4.5102	4.3076	0.0078	0.0071

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	522.427	353.528	398.956	1119.26
mape	0.069	0.0454	0.0512	0.1583
mse	966294	422332	656723	3.17316e+06
total_time	4.3522	3.3333	0.0075	0.0069

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	478.362	361.033	602.926	1340.95
mape	0.0622	0.046	0.0787	0.17
mse	805039	441118	1.61572e+06	6.04619e+06
total_time	5.5485	3.9947	0.0076	0.0072

Plot:

github-actions · 2024-05-03T17:03:15Z

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	4.0435	4.7956	0.0087	0.0048

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	4.3614	3.0167	0.0057	0.0046

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	142.394	196.363	269.23	1331.02
mape	0.0203	0.0234	0.0304	0.1692
mse	63464.7	123119	213677	4.68961e+06
total_time	4.3663	6.228	0.008	0.007

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	522.427	353.528	398.956	1119.26
mape	0.069	0.0454	0.0512	0.1583
mse	966294	422332	656723	3.17316e+06
total_time	7.7752	7.8659	0.0076	0.0071

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	478.362	361.033	602.926	1340.95
mape	0.0622	0.046	0.0787	0.17
mse	805039	441118	1.61572e+06	6.04619e+06
total_time	9.1546	9.0867	0.0076	0.0072

Plot:

github-actions · 2024-05-03T17:53:22Z

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	19.9748	18.7874	0.0083	0.0046

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	22.7074	21.5984	0.0057	0.0047

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	142.394	196.363	269.23	1331.02
mape	0.0203	0.0234	0.0304	0.1692
mse	63464.7	123119	213677	4.68961e+06
total_time	19.9691	20.8946	0.0077	0.0067

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	522.427	353.528	398.956	1119.26
mape	0.069	0.0454	0.0512	0.1583
mse	966295	422332	656723	3.17316e+06
total_time	21.5041	23.2855	0.0074	0.0069

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	478.362	361.033	602.926	1340.95
mape	0.0622	0.046	0.0787	0.17
mse	805039	441118	1.61572e+06	6.04619e+06
total_time	26.2574	22.4733	0.0075	0.0069

Plot:

github-actions · 2024-05-03T21:05:27Z

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	14.384	12.3572	0.0086	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	4040.22	4110.79	5928.17	18859.2
total_time	11.7756	15.404	0.0058	0.0048

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	142.394	196.363	269.23	1331.02
mape	0.0203	0.0234	0.0304	0.1692
mse	63464.7	123119	213677	4.68961e+06
total_time	7.0704	8.0082	0.008	0.0069

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	522.427	353.528	398.956	1119.26
mape	0.069	0.0454	0.0512	0.1583
mse	966295	422332	656723	3.17316e+06
total_time	11.1466	15.1023	0.0075	0.0069

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	478.362	361.033	602.926	1340.95
mape	0.0622	0.046	0.0787	0.17
mse	805039	441118	1.61572e+06	6.04619e+06
total_time	24.4572	21.5935	0.0077	0.0069

Plot:

github-actions · 2024-05-06T17:03:47Z

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	213.935	199.132	2571.33	10604.2
total_time	3.314	3.1755	0.009	0.0049

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	4040.22	4110.79	5928.17	18859.2
total_time	2.9884	2.3238	0.0057	0.0048

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	142.394	196.363	269.23	1331.02
mape	0.0203	0.0234	0.0304	0.1692
mse	63464.8	123119	213677	4.68961e+06
total_time	2.8565	3.272	0.008	0.0072

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	522.427	353.528	398.956	1119.26
mape	0.069	0.0454	0.0512	0.1583
mse	966294	422332	656723	3.17316e+06
total_time	2.9044	3.2741	0.0077	0.007

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	478.362	361.033	602.926	1340.95
mape	0.0622	0.046	0.0787	0.17
mse	805039	441118	1.61572e+06	6.04619e+06
total_time	4.3484	3.4888	0.0079	0.0075

Plot:

AzulGarza

thanks @marcopeix!

i left a qq. also, could you please update mint.json including the new notebooks?
thanks!

AzulGarza · 2024-05-06T20:22:08Z

.github/workflows/ci.yaml

@@ -65,7 +65,7 @@ jobs:
    strategy:
      fail-fast: false
      matrix:
-        os: [ubuntu-latest, macos-latest, windows-latest]
+        os: [ubuntu-latest, macos-13, windows-latest]


is there something in particular that breaks macos-latest? perhaps we can fix it in the notebooks, since the rest of the docu seems to work well with macos-latest.

On local, everything works fine, but in the CI, I encountered this error:

RuntimeError: MPS backend out of memory (MPS allocated: 0 bytes, other allocations: 0 bytes, max allowed: 7.93 GB). Tried to allocate 480.00 KB on private pool. Use PYTORCH_MPS_HIGH_WATERMARK_RATIO=0.0 to disable upper limit for memory allocations (may cause system failure).

I reduced the batch size and val_batch_size, but that didn't fix the CI. José suggested I change it to macos-13.

github-actions · 2024-05-06T20:41:10Z

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	10.8908	6.1772	0.0088	0.0049

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	7.9949	10.2973	0.0059	0.005

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	142.394	196.363	269.23	1331.02
mape	0.0203	0.0234	0.0304	0.1692
mse	63464.7	123119	213677	4.68961e+06
total_time	10.3165	10.4547	0.008	0.007

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	522.427	353.528	398.956	1119.26
mape	0.069	0.0454	0.0512	0.1583
mse	966294	422332	656723	3.17316e+06
total_time	12.5084	13.5082	0.0075	0.007

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	478.362	361.033	602.926	1340.95
mape	0.0622	0.046	0.0787	0.17
mse	805039	441118	1.61572e+06	6.04619e+06
total_time	11.4124	20.4367	0.0076	0.007

Plot:

Electricity demand use case

caf7ffd

marcopeix requested review from mergenthaler and AzulGarza April 29, 2024 20:00

marcopeix added the documentation label Apr 29, 2024

Merge branch 'main' into feature/electricty_demand_example

d89107f

marcopeix marked this pull request as ready for review April 29, 2024 20:34

marcopeix added 2 commits April 30, 2024 09:40

Merge branch 'main' into feature/electricty_demand_example

8349715

Use url for data, better intro

ae3fbe7

Clean notebook

4834fb0

marcopeix added 2 commits May 1, 2024 11:17

Merge remote-tracking branch 'origin/main' into feature/electricty_de…

a15e328

…mand_example

Add nf as dev dependency and use new doc structure

949998c

Fix conflicts

00cf7e9

marcopeix added 2 commits May 2, 2024 10:29

Merge branch 'main' into feature/electricty_demand_example

e5b0639

Fix typo in notebook

f434614

AzulGarza reviewed May 2, 2024

View reviewed changes

nbs/docs/5_use_cases/3_electricity_demand.ipynb Outdated Show resolved Hide resolved

nbs/docs/5_use_cases/3_electricity_demand.ipynb Outdated Show resolved Hide resolved

AzulGarza self-requested a review May 2, 2024 16:58

AzulGarza requested changes May 2, 2024

View reviewed changes

mergenthaler requested a review from AzulGarza May 2, 2024 21:58

Fix conflicts

ff3882d

Reuse dataframes, fix TimeGPT pipeline for good performance

4b64d6c

Reduce batch size for memory issue in CI

0107c2a

Use macos-13 in CI

536f593

Move notebook to right folder

a10430b

marcopeix added 2 commits May 6, 2024 09:17

Merge branch 'main' into feature/electricty_demand_example

b598fb9

Add intermittent demand use case

24de108

AzulGarza requested changes May 6, 2024

View reviewed changes

marcopeix added 2 commits May 6, 2024 16:36

Merge branch 'main' into feature/electricty_demand_example

940ded7

Add notebooks to mint.json

ecf3aba

AzulGarza approved these changes May 6, 2024

View reviewed changes

AzulGarza merged commit e1d2e43 into main May 6, 2024
13 of 14 checks passed

AzulGarza deleted the feature/electricty_demand_example branch May 6, 2024 23:37

mergenthaler pushed a commit that referenced this pull request May 8, 2024

Electricity demand use case (#321)

614805f

Electricity demand use case #321

Electricity demand use case #321

Conversation

marcopeix commented Apr 29, 2024 • edited Loading

review-notebook-app bot commented Apr 29, 2024

github-actions bot commented Apr 29, 2024

Experiment 1: air-passengers

Description:

Results:

Plot:

Experiment 2: air-passengers

Description:

Results:

Plot:

Experiment 3: electricity-multiple-series

Description:

Results:

Plot:

Experiment 4: electricity-multiple-series

Description:

Results:

Plot:

Experiment 5: electricity-multiple-series

Description:

Results:

Plot:

github-actions bot commented Apr 29, 2024

Experiment 1: air-passengers

Description:

Results:

Plot:

Experiment 2: air-passengers

Description:

Results:

Plot:

Experiment 3: electricity-multiple-series

Description:

Results:

Plot:

Experiment 4: electricity-multiple-series

Description:

Results:

Plot:

Experiment 5: electricity-multiple-series

Description:

Results:

Plot:

mergenthaler commented Apr 29, 2024

github-actions bot commented Apr 30, 2024

Experiment 1: air-passengers

Description:

Results:

Plot:

Experiment 2: air-passengers

Description:

Results:

Plot:

Experiment 3: electricity-multiple-series

Description:

Results:

Plot:

Experiment 4: electricity-multiple-series

Description:

Results:

Plot:

Experiment 5: electricity-multiple-series

Description:

Results:

Plot:

github-actions bot commented Apr 30, 2024

Experiment 1: air-passengers

Description:

Results:

Plot:

Experiment 2: air-passengers

Description:

Results:

Plot:

Experiment 3: electricity-multiple-series

Description:

marcopeix commented Apr 29, 2024 •

edited

Loading