Skip to content

Data Type

Note

This document has been machine translated.

column_def

Represents the column information of a dataset. It is represented as a JSON Object with the following entries

  • column_name: string -- column name.
  • data_type: string -- data type.
  • nullable: boolean -- nullable flag. If true, null values are allowed. If false, null values are not allowed and an error is returned. This item can be omitted in the request data. If omitted, it is false.
  • attnum: integer -- Column number in the database. This item may be deprecated in the future. Do not use it. This item cannot be specified in request data.

dataset_info

Represents a dataset of information. It is represented as a JSON Object with the following items.

  • id: integer -- ID on the internal management table. This item may be obsolete in the future. Do not use it.
  • session_id: integer -- session ID. This item may be obsolete in the future. Do not use.
  • usename: string -- Database role name. This field may become obsolete. Do not use.
  • ddc_type: string -- type. This item may be obsolete in the future. Do not use.
  • short_form: string -- dataset name.
  • long_form: string -- long form dataset name. This item may be obsolete in the future. Please do not use it.
  • realname: string -- Name of a table in the database. This field may become obsolete. Do not use.
  • created_at: object -- The date and time the dataset was created. This field may be obsolete in the future. Do not use.
  • committed_at: object -- Commit date and time of the dataset. This item may be obsolete in the future. Do not use.
  • prepared_activity_id: integer -- Operation information for the dataset. This item may be obsolete in the future. Do not use.

citation_info

Represents information on a Citation. It is represented as a JSON Object with the following items

  • id: integer -- The ID on the internal management table. This item may be obsolete in the future. Do not use it.
  • short_form: string -- Citation name.
  • long_form: string -- long form citation name.
  • dataset_name: string -- dataset name.
  • dataset_realname: string -- Name of the dataset table in the database. This item may be obsolete in the future. Do not use it.
  • transaction_time: object -- The date and time when the citation was issued.
  • created_at: object -- The date and time the citation was created in the database. This item may be obsolete in the future. Do not use.

upstream_info

Represents upstream information of a model store. It is represented as a JSON Object with the following items

  • id: integer -- ID on internal management table.
  • model_store_ddc_id: string -- The old-style model store ID on the internal management table. This item may be obsolete in the future. Do not use.
  • model_store_dataset_id: integer -- Model store ID on internal control table. This item may be obsolete in the future. Do not use.
  • upstream_endpoint: string -- The endpoint of the upstream node.
  • upstream_apikey: string -- API key to connect to upstream node.
  • upstream_secret: string -- The secret to connect to the upstream node.
  • upstream_model_store_ddc: string -- The model store name of the upstream node.
  • alias: string -- alias name.
  • description: string -- description.
  • enabled: boolean -- Flag indicating enable/disable.
  • created_at: object -- Date and time the upstream information was created in the database. This item may be obsolete in the future. Do not use.
  • updated_at: object -- The date and time the upstream information was last updated in the database. This field may become obsolete in the future. Do not use.

model_info

Represents model information. It is represented as a JSON Object with the following items.

  • start_datetime: object -- the start date and time represented by the model. This item may be obsolete in the future. Do not use it.
  • end_datetime: object -- the end date and time represented by the model. This item may be deprecated in the future. Do not use.
  • location: string -- The spatial range represented by the model. This field may be obsolete. Do not use.
  • tenant_id: string -- The tenant ID of the xData Edge instance for which the model was created.
  • model_id: string -- The ID that uniquely identifies the model.
  • model_kind: string -- The model type.
  • model_description: string -- A description of the model.
  • model_meta: object -- Meta information about the model.
  • model_state: integer -- A classification number representing the state of the model. It is used to distinguish between states in a federated learning cycle.
    • 10: The learned model.
    • 20: Model fed back from downstream nodes.
    • 30: Aggregated models.
    • 40: model forwarded from an upstream node.
  • model_hash: string -- hash value of the model.
  • model_size: integer -- Size of the model as binary data.
  • round: integer -- number of rounds of coalition training. Currently not used.
  • hop_count: integer -- number of hops in model transfer between nodes. Currently not used.
  • model_store_ddc: string -- The model store name where this model is stored.

condition

Represents the conditions for retrieving various types of information, such as records in a data set. It is expressed in JSON Logic notation. However, the operators available are different from those in JSON Logic.

  • https://jsonlogic.com/

Example Description

{
    "and": [
        { ">=": [ "age", 20 ] },
        { "<" : [ "age", 30 ] }
    ]
}

This example represents the condition that the value of the “age” column is greater than or equal to 20 and less than or equal to 30.

Expressions

A conditional expression is a JSON object that can be written as follows

{
    "operator": [ "operand1", "operand2", ... ]
}
  • A JSON object has only one key. The name of the key determines the type of operation.
  • The value corresponding to the key is a JSON array. Each element of the array is a parameter (operand) of the operation. The number and type of parameters depends on the operator.

Column Names

The column names in the dataset are described by JSON strings.

Values

Numeric values, strings, boolean values, and other immediate values are described by JSON values. However, a string that matches the column name of the dataset is interpreted as a column name.

Operators

The following operators are available.

Operator Description Number of parameters
= A and B are equal 2
<> A and B are not equal 2
> A is greater than B 2
>= A is greater than or equal to B 2
< A is less than B 2
<= A is less than or equal to B 2
`! `` the negation of the parameter 1
and all parameters are true optional
or any parameter is true optional
in A is equal to any element of B 2 (B is a JSON array)
+ the union of all parameters optional
* the product of all parameters optional
- A - B 2
/ A / B 2