How to estimate the size of query results

Question

i have 3 tables
table_h : 10,000 records and 50columns
table_l : 13,000 records and 75 columns
table_ld : 25,000 records and 120 columns
i wrote a query to fetch data based on the 3 tables, and fetching 25,00 records.
ie ( 5 coulms from table_h, 12 coulmns from table_l and 120 columns from table_ld)
how i have to get the size of this query results.

Answer 1

What "size" are we talking about here? There are many different ways to interpret that-- it could be the size of the memory buffer that would be allocated on the client. It could be the size that would be required to store the result in a table in the database. It could be something else. But the answer may differ substantially.
Regardless, it's probably going to be something like (number of rows) * (sum of sizes of each column). But the size required for each column is a bit variable
- Client applications, in general, are going to allocate a memory buffer based on the maximum size of the column (LOB and LONG columns excepted). So if you declare a VARCHAR2(4000) but only store 1 byte in that column, most client applications are going to allocate 4k for that column for each row. It might be even worse than that-- if the client application happens to treat character data as Unicode, it probably uses the UTF-16 encoding in memory, which means allocating 2 bytes for every byte in the database (assuming the database character set supports single-byte characters primarily).
- The Oracle database, on the other hand, is not going to allocate more disk than is actually required to encode variable width data. So the 1 byte column only requires 1 byte of storage (plus a length). That's great for storage efficiency, but that makes computing the actual size a bit more complicated because you now need to know the actual size of data stored in each row (or at least the statistical distribution) rather than the table definition.
Justin