My goal is to develop a workflow that can predict daily sales volumes for the current month, starting from the day after the current date until the end of the month (for example, if today is June 16th, I would like to predict sales from the 17th to the 30th of June). I have historical data starting from January 1st, 2024. It is essential to include seasonality analysis in the model, as I manage a physical store and must account for variations related to national holidays and Sundays.
Hello @Lorenzo_Marinoz
The question in this challenge would be what do you want to do (?)…
You don’t provide previous year historical data; some statistical data is needed aiming to ‘predict’ seasonality. Is this sample all your available data?
Otherwise you just can extrapolate (average daily sales in the month applied to remaining working days…) To do so, business background and knowledge would be the only tool for fine tuning of the prediction.
I can offer you to take a look into the following workflow . Provided data in this workflow, covered one year monthly sales for two products. In the first ‘Column Expressions’ node I can predict seasonality of sales by calculating the differential percentage from previous month. So plotting the differentials gives you a trend about ‘sales seasonality’ .
If your historical data increases (more years), you can illustrate statistics about seasonality (mean differential, standard deviation…); with this type of data you would start to think about time series analysis.
As @MartinDDDD commented, please share what you’ve tried so far.