Stata 14.0 SE on Mac - Data contais 700,000 obervations

Hi. I am trying to reshape a panel of US county population data organised by county and year (data post below)
(fips) is the unique county indicator, (year) is the year indicator and each fips/year has 19 age categories (agegroup).
The seven population variables are currenntly orgaised as: total (tot_pop), total male (tot_male), total females (tot_female)...etc.

What I would like is one observation per fips/year for each combination of age-braket/population-variables. So a total of 133 variables.
i.e. population aged 85+ (a_85_tot_pop),female white population aged 75-80 (a_75_80_wa_female), etc.

I am confused about how to use the reshape commands to achieve this task as it seems l likely have to take my data to wide view, and then back to long view but not sure where to begin here.

In reshape command: i(varname) would include fips s_id c_id year. I'm not sure how to go about identifying the j(varname). (agegroup) is currently string data, while (agegrp) is a numerical indicator for (agegroup)
fips s_id c_id year agegrp agegroup tot_pop tot_male tot_female wa_male wa_female ba_male ba_female
2180 2 180 2000 17 80-84 58 27 31 3 4 0 0
2180 2 180 2001 17 80-84 58 27 31 3 4 0 0
32027 32 27 2002 18 85+ 58 24 34 24 34 0 0
2185 2 185 2003 16 75-80 58 18 40 0 2 0 0
Any help is greatly appreciated,
Thanks,
Mike