Nixtla's Open Source Time Series Ecosystem.

Nixtla Community

a_updated.csv

*Date:* March 28
*Company*: Georgia Institute of Technology
*Contact:* Sarkari, Zenobia &lt;<mailto:zsarkari3@gatech.edu|zsarkari3@gatech.edu>&gt;
_*Hi Team,*_
_*I am trying to familiarize myself with TimeGPT, however, am struggling. Seems to be a different error each time I fix the current one.*_
_*I attached the small csv file I am using as well as the script. Not sure what I am doing wrong.*_

_*CSV attached*_
_*Code:*_ 
_*import pandas as pd*_
_*from nixtlats import TimeGPT*_
_*timegpt = TimeGPT(*_
   _*token='8MzlyaOo7N0uS0Hk71Kt7vHiTg3rQ6CAZ5yF2B15h67dVFK4udDjilMx1fZwnLkgwEy8vwFle9ne7UH4n6jDSn9kE5JfbalB7GWszQ8AJtieByjMukHEp6MhqmfyWH8C6XP2GZHg9ncgE5ZtkYVlTc5ZXP97wqpIDDCUZ5vgVH35N0M7HCR56IbexLZZJhuMDAshv3lXpl5Uy7A2q0qsYOGo1yAXqcvGgnHVcgDbMAz6tRBRRlecdatfeJlN0fVt'*_
_*)*_
 
_*df = pd.read_csv(r'C:\Users\Zenobia\Documents\a_updated.csv')*_
 
_*# # Convert 'PercentChange' column to numeric, coerce errors to NaN*_
_*# df['PercentChange'] = <http://pd.to|pd.to>_numeric(df['PercentChange'], errors='coerce')*_
 
_*# # Drop rows with NaN values in 'PercentChange' column*_
_*# df.dropna(subset=['PercentChange'], inplace=True)*_
_*df.head()*_
 
_*# Forecast*_
_*timegpt.plot(df, time_col='Timestamp', target_col='Close') # Good for report*_
 
_*timegpt_fcst_df = timegpt.forecast(df=df, h=12, freq='MS', time_col='Timestamp', target_col='Close')*_
_*timegpt_fcst_df.head() # No error here*_

El correo queda en la tabla? no hay nada abajo que diga cual es el error que tienen?

porque lo ultimo que veo dice :

_*timegpt_fcst_df.head() # No error here*_

y de hecho estoy viendo en el admin que todos sus calls tienen codigo 200, osea funcionaron jaja. No dice nada mas el correo?

image.png

Hello XXX,

Thanks for trying TimeGPT. Can you describe the errors you are encountering? Which function is failing? We can't see the error in your previous email, and all the API calls processed with your email were successful.

<@U024AQ5STUK> hola, ya nos dieron respuesta:
Hi Diana,
I ended up getting the previous draft to work! Thank you!
However, did the syntax change? I noticed I am having import for the project as well different documentation. It does seem to even import on my end:

They probably need to upgrade the version, pip install is not enough. They need to add -U after nixtlats

Is <@U06MW6C1QM7> going to send this one?