I have a set of data organized as the following table:
Day 1
CIty Temperature Latitude
none [c] [deg]
NYC 10 60
NJ 10 60
Washington 20 50
San Francisco 30 40
Day 2
....
Day 365.
I hope I can store them as a hdf5 file, and use HDFView for example to
plot the latitude .vs. temperature curve for a specific day. and the
column should be expandable, for example I may need one more column on
"populations of each city"
currently I can store pure data as a table, and the city name part as
a new "string type table" with same row as the data table. The
name/meaning of each column, "latitude" and "temperature" and their
"Unit" as other "string type tables", This is quite separated. Seems
not a good scheme. Since I need the expandability, I can not use a
compound datatype to store each line of the above table.
Can anyone shed some like on this problem ?
Thanks.
Forest.