When you retrieve data from LUSID you can typically filter the result set to qualify the data returned.

For example, the following call to the ListInstruments API applies a filter expression of name startswith 'App' to retrieve just those instruments with 'App…'-like names:

curl -X GET "https://<your-domain>" 
  -H "Authorization: Bearer <your-API-access-token>"

Note: There is a maximum URL length of 2048 characters that may impact the complexity of any filter expression you submit via the LUSID API. This restriction does not apply if you filter LUSID data using Luminesce.

As you can see from the table below, the slowest filter expressions are those that join multiple expressions together using the or operator. Note that other factors less easy to tabulate may also affect performance, such as the size of your LUSID dataset and the number of filter expressions you choose to join using either and or or.

SpeedEntity attribute to filter onFilter operatorExample
FasterFields and nested fieldseqname eq 'BP'
 Identifiersidentifiers['Isin'] eq 'GB0007980591'
SHKs (portfolios only)subholdingkeys[Transaction/Demo/Strategy] eq 'Income'
Single-value propertiesproperties[Instrument/Ibor/Industry] eq 'Energy'
Custom data fields (custom entities only)fields['Address'] eq '1 Main Street'
Multi-value propertiesproperties[Instrument/Ibor/Industry] all (~ eq 'Energy')
Relationships (supported entities only)relationships any ( eq 'BP')
As above, in the same ordergt
name startswith 'BP'
As above, in the same orderexists
not exists
not in
name neq 'BP'
As above, in the same orderandname eq 'BP' and domCcy eq 'GBP'
SlowerAs above, in the same orderorname eq 'BP' or domCcy eq 'GBP'