postgresql - Group records with multiple sets in sql -

i have optimize select query less no. of records when have same data different dates. need sort rows date , should combine rows single until finds different column value. typically data below.

date       c_val 1/1/2016    200 2/1/2016    200 3/1/2016    300 4/1/2016    300 5/1/2016    300 6/1/2016    200 7/1/2016    200 

then output should follows.

start_date  end_date    c_val 1/1/2016    2/1/2016    200 3/1/2016    5/1/2016    300 6/1/2016    7/1/2016    200 

the query followed this:

select min(date) start_date, max(date) end_date, c_val t_ord group c_val; 

but returning 2 records grouping c_val. think need additional on order , break when finds new value. there feature available in postgres?

you can use difference of row numbers approach classify consecutive rows (ordered date) same c_val 1 group , start on new group when new value encountered. after done, min , max date of each group per c_val.

select min(date) startdate,max(date) enddate,c_val (select c_val,date,row_number() over(order date)                         -row_number() over(partition c_val order date) grp       t_ord      ) t group c_val,grp;