I'm not quite sure I fully understand the trouble you're having, but I'm game to take a stab at it, even if it's a SQL/PostgreSQL question on a Linux list. (grins) If I understand correctly, you want to be able to 1) give meaningful names to columns that would otherwise have their names defaulted (such as from aggregate functions), and 2) make it easier to read by including the headers with each value You can rename columns using the "AS" keyword if that would help: select grouping_field, round(avg(sug),1) as Rounded_Average, count(*) as Count_of_Records from some_table group by That can be combined with the "\x" command to get the headers on each row with a record-separator. This should produce output something like -[ RECORD 1 ]- grouping_field | alpha Rounded_Average | 3.1 Count_of_Records | 3 -[ RECORD 2 ]- grouping_field | delta Rounded_Average | 6.2 Count_of_Records | 4 -[ RECORD 3 ]- grouping_field | gamma Rounded_Average | 2.5 Count_of_Records | 1 If that doesn't get you what you want, if you can try and describe more clearly what you want, I can take another stab. Hope this helps, -Tim On May 5, 2014, Jude DaShiell wrote: > I wrote a query in sql using psql for an interactive environment in > console mode. I can get a couple different kinds of output neither > of which would be useful for production level work. The column > names from which data was taken in a table would be informative if > those were to show up to the left of the values but that's a > different kind of output and what is also happening when I use an > aggregate function say round(avg(sug),1) is that the function shows > with its result on output. That particular formula was used to > calculate average blood sugar for a set of data. I can use \t and > turn off all tuples and then all I read are the rounded averages > again with no explanatory text. I understand all of this since I > wrote the query but someone else who doesn't know what Linux is > never mind what postgresql is needing to look at this data without > assistance is going to have a hard road to travel. I was reading up > on comment (7) but unless those \D directives can operate like \t > and show column names and comments written for aggregate functions > that won't be a workable solution. > > > > jude <jdashiel@xxxxxxxxxxxxxx> > > _______________________________________________ > Blinux-list mailing list > Blinux-list@xxxxxxxxxx > https://www.redhat.com/mailman/listinfo/blinux-list _______________________________________________ Blinux-list mailing list Blinux-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/blinux-list