Closed alitrack closed 4 years ago
Yes, and in csvq, STDEV can be used.
The names of the functions depends on the system. What do you want for what reason?
I know you can use any name, just want to say, in other popular system, stddev used.
I have no special reason, in fact, it is just a hobby to test csvq, and I did not some comparation with q, similiar app developed with python, packed with pyinstaller, SQL engine is SQLite.
I compare csvq with q, you app is about double speed compare to q, and easy to use.
If you want csvq to get more popular, had better offer a popular SQL syntax, so if new guys use csvq, they do not need learn a new SQL.
that's why recommend base on the MYSQL, PostgreSQL or at least SQLite.
another reason, If I analysis my csv sample with csvq, and then I need big csv with other database, I have go rewrite the SQL, it's not a good idea.
I hope you can understand my English.
OK, I understand. But for example, SQL Server, one of the most used DBMS in the world, uses the function name STDEV. It's better to use standard SQL syntax, but I think it's difficult to decide what is widely used for function names.
Why not just support both function names?
OK, I understand. But for example, SQL Server, one of the most used DBMS in the world, uses the function name STDEV. It's better to use standard SQL syntax, but I think it's difficult to decide what is widely used for function names.
OK, just fine, I seldom use SQL Server now.
Why not just support both function names?
also a good idea.
Having multiple functions that behave the same in one system can confuse users. If it was a solution without any disadvantages, then some existing major systems would have adopted that approach.
Multiple function names might initially confuse some users, but documenting the aliases should reduce this confusion.
Assuming users read the documentation, the difference in function names will not be a problem.
There are differences in function names and other details between any system, and even the SQL syntax differ from system to system. Csvq is not a compatibility program for a specific system. I thought these differences were just trivial issues that users could easily look up and use as needed, but is it really that important?
In most Database and Statistics, we use stddev instead of stdev