The last day () number belongs months EOMONTH inquiry date data;
YEAR (COL) year of the date of the output data, month (col) output date data of the month;
When the difference in the query data exist of: in a first query, then the query only once conditions. The later exists is to run a select statement, check out all the data in the operating conditions statement, multiple queries, the number of queries that table.length;
Where the difference with HAVING: having and where the screening are used, and where the group HAVING screening screening record, the polymerization can be followed by further HAVING functions, where not;
row_number window function: given a unique order data, row_number () over (conditions) as an alias;
case when then else end: Select the condition statement;
cross / full join: includes not only a symbolic link table rows match, further comprising two connecting all records in the table;
left / right (outer) join: both matching symbolic link connection table row, all rows include the left (right) of the connection table;
dateadd (); add or subtract a specified time interval to a date, DATEADD (datepart, number, date): date parameter is valid date expression. number is the number of intervals you want to add; for a time in the future, this number is a positive number, for the last time, this number is negative.
DATEDIFF () function returns the number of days between two dates: DATEDIFF (datepart, startdate, enddate), startdate argument is legitimate and enddate date expression.
Add back left join on condition with the added distinction conditions where: increasing the back where the condition after the temporary table generating, filtering. But on condition that the conditions of use in generating a temporary table, regardless of whether it is on the condition is true, it will return the records to the left of the table.