LokqlDx‐Loading data
Lokqldx holds data in tables internally. The current collection of tables is called the context.
When you first create or load a workspace, the context will be empty and you'll need to re-execute the relevant commands to populate the context.
Tables can be loaded from a number of sources:
- 'inline' data using the addtable command
- external files (csv,excel,json,parquet,text)
- results of queries
- Azure ApplicationInsights
The addtable command can be used to quickly add small tables from inline data.
.addtable sampleData
A | B
1 | 2
3 | 4
External files can be loaded using the .load path-to-file tablename command. The file format is determined by the extension. If no table-name is specified, the file-name (without the extension) is used to name the table. I.e. the command .load C:\mydata\ages.csv will create a table called ages. By default, the .load command will skip the reload of data if the target table already exists but the -f flag can be used to force a reload if, for example, you want to fetch updated data from the file.
Dragging and dropping a file with a recognised format into the query editor will automatically insert a .load <file-path> command into the editor. The command can then be run to load the file.
Csv files are a common way to store data and are supported. The csv loader has a number of configurable options which are controlled via the .set command...
It's common for csv files to
By default only the first sheet is loaded by the .load command.
The loadexcel command cane be used to load all worksheets (one table per worksheet)
The results of a previous query can be stored back into the context using the materialize command.
.load dataset.csv
dataset
| extend Category=case(.....)
.materialize categorised
categorised | count
categorised | summarize count() by Category
see....