Operators overview
An operator
is a code module that provides a capability to perform an operation
on a discoveryspace
. For example the RandomWalk
operator provides the capability to perform a random walk operation
on a discoveryspace
.
The pages in this section (see left hand nav menu) give details about some of the operators available in ado
: what they are for, what they do and how to use them.
Info
In addition, the examples section contains worked examples of using some of these operators.
operator
types¶
Operators are grouped into the following types:
- explore: sample and measure entities from a
discoveryspace
- characterize: analyse a
discoveryspace
- modify: create a new
discoveryspace
by changing the entityspace or measurementspace of an inputdiscoveryspace
- compare: compare one or more
discoveryspaces
- fuse: create a new
discoveryspace
from a set of inputdiscoveryspaces
This page describes explore operators in more detail as they are the only operators that sample and measure entities.
Listing the available operators¶
The following CLI command will list the available operators
ado get operators
OPERATOR TYPE
1 detect_anomalous_series characterize
0 profile characterize
2 random_walk explore
3 ray_tune explore
9 export_to_llm_lakehouse export
8 integrate_and_export_to_llm_lakehouse_format export
4 add_experiment modify
7 generate_representative_subspace modify
5 learning_split modify
6 rifferla modify
Using operators¶
Using an operator involves the following steps:
- Find the input parameters for the operator:
ado template operation --operator-name $OPERATOR_NAME
- Write an operation YAML for applying the operator.
- This involves setting specific values for its input parameters.
- Create the operation:
ado create operation -f $YAML
- Retrieve the results of the operation:
ado show related $OPERATION_ID
- in addition
ado show entities $OPERATION_ID
for explore operations
These steps are covered in detail in operations.