Some relational database management systems (RDBMSs) have a PERCENTILE_CONT() function that calculates a percentile based on a continuous distribution across a range of column values.
We specify the percentile to use when we call the function.
Depending on the RDBMS, PERCENTILE_CONT() can be used as a window function or an aggregate function, or as both.
Many relational database management systems (RDBMSs) provide a window function called PERCENT_RANK() that returns the relative rank of a row within a group of rows.
The relative rank is expressed as a value between 0 and 1.
Some SQL databases have a window function called NTH_VALUE() that allows us to get a value from a given row in the window frame, based on the row number.
More specifically, the function returns the value of a given expression from the from the N-th row of the window frame, where N is a number that we specify when calling the function.
In SQL databases, LAST_VALUE() is a window function that returns the last value in an ordered set of values. It allows us to get a value from the last row of a query result set or partition.
You may need to explicitly set the window frame if you want LAST_VALUE() to return the actual last value from the partition or result set. That’s because in many/most DBMSs, the default window frame ends with the current row.
If you’re getting an error that reads “function mode() does not exist” in PostgreSQL, it could be that you’re using the wrong syntax with this function.
The mode() function requires a WITHIN GROUP clause, and we can get the above error if we remove that clause.
In this case, we can fix the error by adding a valid WITHIN GROUP clause.
In SQL databases, the FIRST_VALUE() function is a window function that returns the first value in an ordered set of values.
The exact syntax will probably depend on your DBMS, but it will usually require an OVER clause to determine how the rowset is partitioned and ordered before the window function is applied.
Many relational database management systems (RDBMSs) provide a DENSE_RANK() function that we can use in our SQL queries. The SQL DENSE_RANK() function is a window function that returns the rank of the current row within its partition, without gaps.
The “without gaps” part is what distinguishes the DENSE_RANK() function from the RANK() function.
DENSE_RANK() returns contiguous rank numbers whenever there are ties, whereas RANK() will leave a gap between the tie and the next rank, resulting in noncontiguous rank numbers.
In SQL databases, the RANK() function is a window function that returns the rank of the current row within its partition, with gaps.
By “with gaps” I mean that if there are any ties for a given rank, there will be a gap between that rank value and the next rank value.
If you don’t want such gaps, use the DENSE_RANK() function instead, as it returns the rank without gaps.
Many relational database management systems (RDBMSs) provide us with a range of window functions.
The CUME_DIST() function is a window function that’s commonly implemented in RDBMSs for the purpose of calculating the cumulative distribution across a data set. In other words, it calculates the relative position of a specified value in a group of values.
In PostgreSQL the last_value() function returns the value from the last row of the current window frame.