https://www.nixtla.io logo
Join Slack
Powered by
This message was deleted.
# general
s
This message was deleted.
y
lol i think i know why . 3.11 is not supported yet.. https://github.com/Nixtla/statsforecast/pull/425
j
Hey. Can you print the types of the fcst_df (
fcst_df.dtypes
)?. I think the level columns may not be numerical
m
Hi @Yaping Lang, thanks for the interestest we are currently in closed beta and are trying to onboard as many users as possible. That means that you probably are getting a 429 error. Something like:
Copy code
{
  "status": 429,
  "data": null,
  "message": "Too many requests",
  "details": "You have reached your request limit, top up your credits to continue using the API",
  "code": "A20",
  "requestID": ""
}
If you haven't yet. Could you please fill out the form on our Website so we can process it? Just click on Submit interest on https://www.nixtla.io/
y
@José Morales, they were float
Copy code
unique_id         object
ds                object
TimeGPT          float64
TimeGPT-lo-90    float64
TimeGPT-lo-80    float64
TimeGPT-hi-80    float64
TimeGPT-hi-90    float64
dtype: object
j
Hmm I think the ds has to be timestamp. Which version of statsforecast are you using?
Also if you can try installing utilsforecast instead (
pip install utilsforecast
) and using:
Copy code
from utilsforecast.plotting import plot_series

plot_series(df, fcst_df, level=[80, 90], max_insample_length=24 * 5)
It'd help us because we're going to replace the statsforecast function with that one and it'd be better to fix the error there if we find one
10 Views
Open in Slack
PreviousNext
Powered by