Primary competition visual

agriBORA Commodity Price Forecasting Challenge

Helping Kenya
€8 250 EUR
18 days left
Data analysis
GIS
Time-series
Forecasting
Nowcasting
604 joined
199 active
Starti
Nov 14, 25
Closei
Dec 27, 25
Reveali
Jan 13, 26
User avatar
wako
Adama science and technology university
how to merge the new datasets to the old one to predict?
Help · 22 Nov 2025, 11:50 · 3

after the week 46 actual price has released , how to merge to the old datasets to retrain my model and for which week the prediction will needed since we have already done for week 48 and 49,on the other hand how will be the previous week prediction, will it be zero or continue with its predicted value ?

Discussion 3 answers
User avatar
Juliuss
Freelance

good point. @AJoel, @Zindi, can we just get additional weekly data merged with the original agribora_maize_prices.csv from your end and then some alert chat? instead of multiple weekly update files?

22 Nov 2025, 23:31
Upvotes 0
User avatar
AJoel
Zindi

Hi @wako and @Juliuss, I have update the weekly dataset. It now has the same structure as "agribora_maize_prices.csv". So a simple merge should work.

## Merge agriBORA_df & df
agriBORA_df = pd.read_csv('agribora_maize_prices.csv')
df_week_46_47 = pd.read_csv('agriBORA_maize_prices_weeks_46_47.csv')
df = pd.concat([agriBORA_df, df], ignore_index=True)
User avatar
Juliuss
Freelance

Thanks alot. That will do