Daily Time Series Forecasting

[agustira97]

Dear all member of Alteryx Community,

I want to do forecasting. I have daily data in 1 month (2019-05) and i want the result of my forecasting is daily, from 2019-06-01 until 2019-06-30. I try to follow tutorial from youtube, but the case usually use monthly forecast. I need help for my case.

Thank you.
*sorry for my bad english

Sample file attached.

[Laurap1228]

Here is the link to pretty good blog on using the ARIMA Model

https://visualbi.com/blogs/self-service-bi/alteryx/forecasting-arima-model-alteryx/ 

[StephaneP]

Hello @agustira97 ,

1st point. The algorithm do not care if your frequency is month, week or day. It only have to be consistent. Always days or always week in your table. Moreover you must not have any missing value. Every period (days here) have to be populated. Data preparation is important.

To learn in detail how to use Dynamic Time series here are some ressources in complement of the one of @Laurap1228 :

• The Alteryx DataScience learning path that have a dedicated focus on the Dynamic Time Series: Data Science Learning path 

• 2 dedicated articles on how to use the 2 DTS algorithm
• How to use ARIMA 
•  How to use ETS 

And if you prefer videos, around 1h with real workflow to download:

• Time series modeling 
• Time Series Analysis 

You also have example of CashFlow Forecasting using DTS in the Office of Finance Starter kit that can be download here: Office of Finance Starter Kit 

And finally the examples in the Designer Help menu:

Hope it helps