label_mapper-1.1.0¶

Represents a mapping from a coordinate value to a label.

Description

A label mapper instance maps inputs to a label. It is used together with regions_selector. The label_mapper returns the label corresponding to given inputs. The regions_selector returns the transform corresponding to this label. This maps inputs (e.g. pixels on a detector) to transforms uniquely.

Schema Definitions ¶

This node must validate against all of the following:

This type is an object with the following properties:
- mapper
  object Required
  A mapping of inputs to labels. In the general case this is a astropy.modeling.core.Model.
  
  It could be a numpy array with the shape of the detector/observation. Pixel values are of type integer or string and represent region labels. Pixels which are not within any region have value no_label.
  
  It could be a dictionary which maps tuples to labels or floating point numbers to labels.
  This node must validate against any of the following:
  
  ../core/ndarray-1.0.0
  transform-1.1.0
  This type is an object with the following properties:
  
  labels
  array
  No length restriction
  
  Items in the array must be any of the following types:
  
  number
  
  array
  No length restriction
  
  Items in the array are restricted to the following types:
  
  number
  
  models
  array
  No length restriction
  
  Items in the array are restricted to the following types:
  
  transform-1.1.0
- inputs
  array
  Names of inputs.
  
  No length restriction
  
  Items in the array are restricted to the following types:
  
  string
  No length restriction
- inputs_mapping
  transform-1.1.0
  mapping
- atol
  number
  absolute tolerance to compare keys in mapper.
- no_label
  object
  Fill in value for missing output.
  This node must validate against any of the following:
  
  number
  
  string
  No length restriction

Examples ¶

Map array indices are to labels.:

!transform/label_mapper-1.1.0
  mapper: !core/ndarray-1.0.0
    [[1, 0, 2],
    [1, 0, 2],
    [1, 0, 2]]

Map numbers dictionary to transforms which return labels.:

!transform/label_mapper-1.1.0
  mapper: !!omap
  - !!omap
    labels: [-1.67833272, -1.9580548, -1.118888]
  - !!omap
    models:
    - !transform/compose-1.1.0
      forward:
      - !transform/remap_axes-1.1.0
        mapping: [1]
      - !transform/shift-1.1.0 {offset: 6.0}
    - !transform/compose-1.1.0
      forward:
      - !transform/remap_axes-1.1.0
        mapping: [1]
      - !transform/shift-1.1.0 {offset: 2.0}
    - !transform/compose-1.1.0
      forward:
      - !transform/remap_axes-1.1.0
        mapping: [1]
      - !transform/shift-1.1.0 {offset: 4.0}
  inputs: [x, y]
  inputs_mapping: !transform/remap_axes-1.1.0
    mapping: [0]
    n_inputs: 2

Map a number within a range of numbers to transforms which return labels.:

!transform/label_mapper-1.1.0
  mapper: !!omap
  - !!omap
    labels:
    - [3.2, 4.1]
    - [2.67, 2.98]
    - [1.95, 2.3]
  - !!omap
    models:
    - !transform/compose-1.1.0
      forward:
      - !transform/remap_axes-1.1.0
        mapping: [1]
      - !transform/shift-1.1.0 {offset: 6.0}
    - !transform/compose-1.1.0
      forward:
      - !transform/remap_axes-1.1.0
        mapping: [1]
      - !transform/shift-1.1.0 {offset: 2.0}
    - !transform/compose-1.1.0
      forward:
      - !transform/remap_axes-1.1.0
        mapping: [1]
      - !transform/shift-1.1.0 {offset: 4.0}
  inputs: [x, y]
  inputs_mapping: !transform/remap_axes-1.1.0
    mapping: [0]
    n_inputs: 2

Original Schema ¶

%YAML 1.1
---
$schema: "http://stsci.edu/schemas/yaml-schema/draft-01"
id: "http://stsci.edu/schemas/asdf/transform/label_mapper-1.1.0"
title: >
  Represents a mapping from a coordinate value to a label.
description: |
  A label mapper instance maps inputs to a label.  It is used together
  with
  [regions_selector](ref:schemas/regions_selector-1.1.0). The
  [label_mapper](ref:schemas/label_mapper-1.1.0)
  returns the label corresponding to given inputs. The
  [regions_selector](ref:schemas/regions_selector-1.1.0)
  returns the transform corresponding to this label. This maps inputs
  (e.g. pixels on a detector) to transforms uniquely.

examples:
  -
    - Map array indices are to labels.

    - |
        !transform/label_mapper-1.1.0
          mapper: !core/ndarray-1.0.0
            [[1, 0, 2],
            [1, 0, 2],
            [1, 0, 2]]

  -
    - Map numbers dictionary to transforms which return labels.

    - |
        !transform/label_mapper-1.1.0
          mapper: !!omap
          - !!omap
            labels: [-1.67833272, -1.9580548, -1.118888]
          - !!omap
            models:
            - !transform/compose-1.1.0
              forward:
              - !transform/remap_axes-1.1.0
                mapping: [1]
              - !transform/shift-1.1.0 {offset: 6.0}
            - !transform/compose-1.1.0
              forward:
              - !transform/remap_axes-1.1.0
                mapping: [1]
              - !transform/shift-1.1.0 {offset: 2.0}
            - !transform/compose-1.1.0
              forward:
              - !transform/remap_axes-1.1.0
                mapping: [1]
              - !transform/shift-1.1.0 {offset: 4.0}
          inputs: [x, y]
          inputs_mapping: !transform/remap_axes-1.1.0
            mapping: [0]
            n_inputs: 2

  -
    - Map a number within a range of numbers to transforms which return labels.

    - |
        !transform/label_mapper-1.1.0
          mapper: !!omap
          - !!omap
            labels:
            - [3.2, 4.1]
            - [2.67, 2.98]
            - [1.95, 2.3]
          - !!omap
            models:
            - !transform/compose-1.1.0
              forward:
              - !transform/remap_axes-1.1.0
                mapping: [1]
              - !transform/shift-1.1.0 {offset: 6.0}
            - !transform/compose-1.1.0
              forward:
              - !transform/remap_axes-1.1.0
                mapping: [1]
              - !transform/shift-1.1.0 {offset: 2.0}
            - !transform/compose-1.1.0
              forward:
              - !transform/remap_axes-1.1.0
                mapping: [1]
              - !transform/shift-1.1.0 {offset: 4.0}
          inputs: [x, y]
          inputs_mapping: !transform/remap_axes-1.1.0
            mapping: [0]
            n_inputs: 2

allOf:
  - $ref: "transform-1.1.0"
  - type: object
    properties:
      mapper:
        description: |
          A mapping of inputs to labels.
          In the general case this is a `astropy.modeling.core.Model`.

          It could be a numpy array with the shape of the detector/observation.
          Pixel values are of type integer or string and represent
          region labels. Pixels which are not within any region have value ``no_label``.

          It could be a dictionary which maps tuples to labels or floating point numbers to labels.

        anyOf:
          - $ref: "../core/ndarray-1.0.0"
          - $ref: "transform-1.1.0"
          - type: object
            properties:
              labels:
                type: array
                items:
                  anyOf:
                    - type: number
                    - type: array
                      items:
                        type: number
                      minLength: 2
                      maxLength: 2
              models:
                type: array
                items:
                  $ref: "transform-1.1.0"

      inputs:
        type: array
        items:
          type: string
        description: |
          Names of inputs.
      inputs_mapping:
        $ref: "transform-1.1.0"
        description: |
          [mapping](ref:schemas/remap_axes-1.1.0)
      atol:
        type: number
        description: |
          absolute tolerance to compare keys in mapper.
      no_label:
        description: |
          Fill in value for missing output.
        anyOf:
          - type: number
          - type: string

    required: [mapper]
...