This site uses different types of cookies, including analytics and functional cookies (its own and from other sites). To change your cookie settings or find out more, click here. If you continue browsing our website, you accept these cookies.
I'm trying to do some time series analysis. Is there I way I can set the number of points that each of my period has? The problem I'm having is that all the options in the for target fild frequencey only has options for hours, days, weeks, etc. My data does not match any of those options. Each period in my data has about 300 points. I would like to set this in tool: Tell tool that my periods have exactly 300 points.
There is an option for Other but when you click it, nothing happens. Still no option to set the number of points in the period.
When you say points are you referring to records in your data? If you are able to attach a clean example that would really help. In general if you are trying to forecast for say more that one "thing" like a region, a product, or something like that you probably want to look at the "TS Forecast Factory" or "TS Model Factory" tools. If you can help me understand what you mean by points then I should be able to help!
Also there is a great time series example already built in Alteryx if you go here: Help > Sample Workflows > Predictive Tool Samples > Predictive Analytics > 15 Time Series Forecasting Sample
I think it's not possible. When you select other and then go the model customization tab, there is nothing explicitly saying frequency. There are several options about the coefficients of the model. I don't know if through modifying the coefficients, I can set the set the frequency. I would need to know exactly what each coefficient means and those details are beyond my current knowledge.
Yes, when I say points I mean records in my data. I have about 300 records per period. All the records are part of the same product, so I cannot model them separately using the TS model factory tool. The point of having all the records together in one single time series model is that there are patterns repeat period by period and I want the model to learn from those patters that repeat (seasonality). If I model them separately, I would lose that information.
The following is a dummy example that captures what I mean. Notice that each period only has 5 values in my dummy example (in my real data I have about 300 rows per period). Notice that there is seasonality. Period by period there is a patterns that repeats.
I wan to model something similar to this, but in this example, I need to tell my model that each period has only 5 points. How do I do this on Alteryx?