Hi guys, I couldn't make the future exogenous valu...
# timegptx
Xinrui Shao
05/14/2024, 12:49 PMHi guys, I couldn't make the future exogenous values work. Every time I pass the X_df, I got the response of 6 attempts, and every attempts fail, in the end it returns 500 internal error.
Copy code
INFO:nixtlats.nixtla_client:Using the following exogenous variables: holiday
INFO:nixtlats.nixtla_client:Calling Forecast Endpoint...
INFO:nixtlats.nixtla_client:Attempt 1 failed...
INFO:nixtlats.nixtla_client:Attempt 2 failed...
INFO:nixtlats.nixtla_client:Attempt 3 failed...
INFO:nixtlats.nixtla_client:Attempt 4 failed...
INFO:nixtlats.nixtla_client:Attempt 5 failed...
INFO:nixtlats.nixtla_client:Attempt 6 failed...Xinrui Shao
05/14/2024, 12:53 PMThis is the code block:`start='2023-06-06 000000'`
forecast_df = nixtla_client.forecast(
df=filtered_df[:start].drop(columns=['ds','flow'])[['unique_id','flow_smooth']+['holiday']].reset_index(),
X_df=filtered_df['2023-06-06 001500':'2023-06-06 040000'].drop(columns=['ds','flow'])[['unique_id']+['holiday']].reset_index(),
h=16,
freq='15T',
finetune_steps=10,
id_col='unique_id',
target_col='flow_smooth',
time_col='ds'
)
Xinrui Shao
05/14/2024, 12:53 PMm
Marco
05/14/2024, 1:13 PMThis is usually due to some latency, and not error in your code! Can you try running it again? 🙂
x
Xinrui Shao
05/14/2024, 1:20 PMI see, I tried to reduce the payload size, it worked! thanks for the help
🙌 1