On Tuesday, May 21, 2013 11:51:09 PM UTC+12, Garba wrote:
> TideMan <mulgor@gmail.com> wrote in message <9dbad8fa-0ebf-4846-8bea-3b3ba71eb91a@googlegroups.com>...
>
> > On Tuesday, May 21, 2013 10:08:09 PM UTC+12, Garba wrote:
>
> > > I have a data set, a daily degree data (366 data points) that I want to convert to hourly values (24*366 = 8784 data points). The raw data obtained are averaged hourly values. So I want to reconvert to the hourly values by generating random hourly values that will represent the daily average values. Any hint on a MATLAB script that will do this?
>
> >
>
> > You need to specify both the distribution and the scale parameter.
>
> > e.g., normal distribution with standard deviation of x.
>
> >
>
> > Then you would do this:
>
> > y=ybar*ones(1,24) + x*randn(366,24);
>
> > y=y(:);
>
> > where ybar is a column vector (366 x 1) of daily means.
>
> Thanks very much it works, but when I average the 24 hourly values they dont equal the daily value
ry=x*randn(366,24);
rybar=mean(ry,2);
y=(ybar-rybar)*ones(1,24) + ry;
> TideMan <mulgor@gmail.com> wrote in message <9dbad8fa-0ebf-4846-8bea-3b3ba71eb91a@googlegroups.com>...
>
> > On Tuesday, May 21, 2013 10:08:09 PM UTC+12, Garba wrote:
>
> > > I have a data set, a daily degree data (366 data points) that I want to convert to hourly values (24*366 = 8784 data points). The raw data obtained are averaged hourly values. So I want to reconvert to the hourly values by generating random hourly values that will represent the daily average values. Any hint on a MATLAB script that will do this?
>
> >
>
> > You need to specify both the distribution and the scale parameter.
>
> > e.g., normal distribution with standard deviation of x.
>
> >
>
> > Then you would do this:
>
> > y=ybar*ones(1,24) + x*randn(366,24);
>
> > y=y(:);
>
> > where ybar is a column vector (366 x 1) of daily means.
>
> Thanks very much it works, but when I average the 24 hourly values they dont equal the daily value
ry=x*randn(366,24);
rybar=mean(ry,2);
y=(ybar-rybar)*ones(1,24) + ry;