Dear all,
I want to build a basic prediction model.
I have a dependent variable that MAY change everyday except Sunday and Monday.
And several independent variableS that change everyday except Saturday and Sunday.
I wish to build a correlation, maybe creating a change factor and accumulating the sum of values of independent variables until I have a change on the dependent. And using the independent values on day+1 and day+2 of the dependent variable.
But the weekends are an issue for me. I wish to avoid those gaps, or create a pattern to work upon them, since the dependent and independents have differents days of weekeds (sunday and monday against saturday and sunday)
I wish to know if there is a way to choose the week days on a correlation, like my Saturday independent variable d+1 wont see the real data from the Sunday data from dependent variable, and my d+2 won't work with the monday dependent variable, because there is no data, orelse will be the same data as Saturday.
I also wish to know if there is a way to count each 0 factors on the dependent variable change factor, like, I wish to know how many 0 (days without change) since the last change. And acumulate the sum of the independent variable upon this result.
I've attached and example to help understand.
Thanks in advance.
Tanai Gonçalves.