creating a histogram

[Martin Imelda Borg]

Hello there:
Q: I would like to create a histogram showing the PROPORTION OF CAR MECHANICS vs the mean no of CAR repairs done per year, PER car mechanic

See picture:

**Repair - if a repair was performed (1) or not (0)
**Totaldoneperyear - total no of repairs performed each year generated from:

bys mechanic repairyear: egen totaldoneperyear=sum(repair)

clear
input float(repair repairdate) str1 mechanic float(repairyear totaldoneperyear)
1 14610 "1" 2000 1
0 15310 "1" 2001 0
1 15745 "1" 2003 1
0 16109 "1" 2004 0
0 16468 "1" 2005 0
1 17867 "2" 2008 1
0 17932 "2" 2009 0
1 18298 "2" 2010 2
1 18303 "2" 2010 2
1 19029 "2" 2012 1
end
format %td repairdate

I've tried this:

histogram totaldoneperyear, freq by(mechanic)

However this generates differ panels for each mechanic. I would like them all in one panel as seen above (in my very amateurish drawing)
Can anyone help?

[George Ford]

drop the by

[Martin Imelda Borg]

Actually found an alternative to

collapse (mean) repair, by (mechanic)
histogram repair, freq

is there another alternative to -collapse- ?
-> this code : generates the average no of repairs done by the mechanic throughout the study period and gives each mechanic ONE row only

if I leave the code as post #1 I only have the ‘total number’ done by the mechanic which if the mechanic does 2 repairs in 1990 as there are 2 rows for 1990, the histogram from post #1 will present this as a total of 4 (as you can see in the data)

just wondering if there is an alternative to -collapse- to generate the average number of repairs done by the mechanic throughout the study period and rom that plotting a histogram with the frequency of mechanics performing average repairs per year

not sure if I made myself clear?

[George Ford]

bys mechanic: g mechanicN = _n
egen repair_m = mean(repair), by(mechanic)
hist repair_m if mechanicN==1, freq