Rolling regressions

[nightwalk]

I would greatly appreciate some help,

First i have some intraday data that i want to input in Eviews, meaning that i have many observations for the same date. For example, i have 60 obs for Dec 22, 2008, 59 obs for Dec 21, 2008 etc. How can i load such kind of data?

When the data are loaded, i want to run rolling linear regressions of the form: ri = a+brm, meaning that i want to estimate the b coefficient for all dates in the sample. How can i do such an estimation?

After the rolling regressions are estimated, is there any way i could get the b coefficients for all dates in a separate consolidated file, sorted by date?

Merry Christmas!