Queries to reduce a set of data pairs

I have a table of values which has the following fields:

AreaElement, Joint, GlobalX, GlobalY, M_sd

This list consists basically in a long list of M_sd values. Each
AreaElement has 4 Joints, each one with GlobalX and GlobalY as it's
coordinates along with a M_sd as it's value. Yet, although different
AreaElements may share common Joints, all M_sd values are independent
of each element. So, for a small example, consider the following
sample:

sample
AreaElem Joint GlobalX GlobalY M_sd
1000 ~1044 6.32 11.91 12.17
1000 ~919 6.32 12.15 10.50
1000 ~1052 6.56 11.91 13.50
1000 ~927 6.56 12.15 12.00
1001 ~1045 6.56 10.20 20.46
1001 ~1046 6.56 10.44 20.32
1001 1733 6.80 10.20 19.38
1001 ~1053 6.80 10.44 19.31
1002 ~1046 6.56 10.44 20.13
1002 ~1047 6.56 10.69 19.39
1002 ~1053 6.80 10.44 19.60
1002 ~1054 6.80 10.69 18.86
/sample

What I want to do is for each AreaElement's Joints which share a
common GlobalX calculate the average of it's M_sd values and it's
GlobalY values and return a list consisting of the following fields:

AreaElem Joint GlobalX Avg(GlobalY) Avg(M_sd)

Is this possible? If so, how is it done?


Thanks in advance,
Rui Maciel