We have an SQL database with several thousend tables. Each table is one single time-series.
We have managed to create a new data set with one table. The import is working just fine and we can chart the data.
We would like to create a data set with dozens of time series. How should we do that? We cannot create hundreds of different SQL connections.
Size:
Color:
By saying "time series", do you mean Named DataSeries? Unfortunately, the Database provider was not designed with multiple time series in mind. Per each table (DataSet) with DTOHLCV data it's possible to add a single Named DataSeries only.
Size:
Color:
Yes, that is correct. Each table is one "Named DataSeries". We hold most of our data in a SQL data base. It would really help us, if you could find a solution.
Size:
Color:
QUOTE:
Each table is one "Named DataSeries".
^^^
Please clarify if the above describes your current situation because previously you said kind of an opposite thing:
QUOTE:
We have managed to create a new data set with one table.
Multiple connections per DataSet are not (and will never be) considered. The idea is to evaluate possibility of supporting multiple Named DataSeries which reside in a
single table.
Size:
Color:
Thanks for the answer.
I just talked with my programmer. Multiple NameDataSeries in a single table should work for us. We would just put each "Data List" into a single table.
We do not need multiple connections per DataSet.
Size:
Color:
Got it, thanks for the clarification. Item added to our backlog for investigation.
Note, however, that having thousands of columns in a database table may slow down the New DataSet Wizard during the DataSet creation step.
Size:
Color:
That's correct.
We have the idea to put together baskets of only 10 to 50 time series in one database table (f.ex. US equity indices, US$ bond indices, ...)
Size:
Color:
Please note that if you're going to put too many symbols (e.g. thousands) in a database table, this may lead to the same slowdown during DataSet creation (while Windows controls are being populated).
Size:
Color:
We have been successful in combing several time-series into one (SQL) database table, that is readable by WL! We can now read these different SQL tables into WL and get one dataset per SQL table.
The issue: "idea is to evaluate possibility of supporting multiple Named DataSeries which reside in a single table" has been solved for us.
Size:
Color:
That's good news to hear. It's not quite clear to me how this overcomes the current limitation of 1 Named DataSeries per DataSet, though. But whatever works for you is fine with me.
Size:
Color: