Query Constraints, Notes on Timestamps, Think before you query
Constraints on a Query
The only constraints on a query are the time range you are selecting and the symbols you are selecting.
Note that all Timestamp in CloudQuant Liberator are in New York Time (US ET). Even if the dataset is for a market on the other side of the world, we will adjust the timestamp to represent, as best we can, the time of arrival in NY ET so that you can easily know when it is available.
Notes on TimeStamps
As such many datasets may contain two or more timestamps, the main timestamp used for the query is the time we expect the data to be available to you. There may be an exchange or execution timestamp which will be before or the same as the main timestamp, there may also be a bar_time if the data is rolled into a bar. IE for minute_bars, the first bar of the day in US Equities market will end at 9:30:59.999999 as such the timestamp will be 9:30:59.999999 but the bar_time will be 9:31:00.000000. You will find this for many ‘calculated’ datasets. The data is complete and available, some calculation time may be taken, and then the bar is available. So for some vendors a bar may have a timestamp of 9:31:05.487282, a bar_time of 9:30:00 and a vendor timestamp of 9:31:00.00000. Our timestamp is to give you the best possible idea of the ACTUAL AVAILABLE TIME OF THE DATA.
Think before you submit the first time
It is very important, when submitting any query, to take a moment to consider the likely size of the request.
For example, if I was to ask for data for AAPL for the month of April from a dataset called daily_bars I can expect to get around 30 rows of data back. But if I submit the same query to a dataset called ‘NBBO’ containing all Bids and Asks I may get an enormous dataset back. And for the Liberator system to execute that request it will need to scan through all bids and offers for all symbols for the whole month looking just for AAPL updates, you can imagine how long that query may take to execute.
So, always be thoughtful about the Queries you submit, especially to unfamiliar datasets. Try a limited date/time range until you start to get some data, then pick a symbol and pull it for a longer and longer time range until you are comfortable.
For advice, you can always reach out to customer_success@cloudquant.com