Selecting Algorithms

Select by Metadata Attribute

Path: flexiconc/algorithms/select_by_metadata_attribute.py

Description:

Selects lines based on whether a specified metadata attribute compares to a given target value. If a list is provided as the target value, membership is tested using equality. For a single numeric value, a comparison operator (==, <, <=, >, >=) can be specified. For strings, only equality (with optional regex matching and case sensitivity) is supported.

Arguments:

Name	Type	Description
metadata_attribute	string	The metadata attribute to filter on.
value	['string', 'number', 'array']	The value to compare against, or a list of acceptable values. When a list is provided, only equality is used.
operator	string	The comparison operator for numeric comparisons. Only allowed for single numeric values. Default is '=='.
regex	boolean	If True, use regex matching for string comparisons (only with equality). Default is False.
case_sensitive	boolean	If True, perform case-sensitive matching for strings. Default is False.
negative	boolean	If True, invert the selection. Default is False.

Show full JSON schema

{
  "type": "object",
  "properties": {
    "metadata_attribute": {
      "type": "string",
      "description": "The metadata attribute to filter on.",
      "x-eval": "dict(enum=list(set(conc.metadata.columns) - {'line_id'}))"
    },
    "value": {
      "type": [
        "string",
        "number",
        "array"
      ],
      "description": "The value to compare against, or a list of acceptable values. When a list is provided, only equality is used."
    },
    "operator": {
      "type": "string",
      "enum": [
        "==",
        "<",
        "<=",
        ">",
        ">="
      ],
      "description": "The comparison operator for numeric comparisons. Only allowed for single numeric values. Default is '=='.",
      "default": "=="
    },
    "regex": {
      "type": "boolean",
      "description": "If True, use regex matching for string comparisons (only with equality). Default is False.",
      "default": false
    },
    "case_sensitive": {
      "type": "boolean",
      "description": "If True, perform case-sensitive matching for strings. Default is False.",
      "default": false
    },
    "negative": {
      "type": "boolean",
      "description": "If True, invert the selection. Default is False.",
      "default": false
    }
  },
  "required": [
    "metadata_attribute",
    "value"
  ]
}

---

Select by Rank

Path: flexiconc/algorithms/select_rank_wrapper.py

Description:

Selects lines based on rank values obtained from the ranking keys in the ordering_result['rank_keys'] of the current node, by default by the first ranking key.

Arguments:

Name	Type	Description
ranking_column	string	The ranking column to use for selection.
comparison_operator	string	The comparison operator to use for the ranking scores.
value	number	The numeric value to compare the ranking scores against.

Show full JSON schema

{
  "type": "object",
  "properties": {
    "ranking_column": {
      "type": "string",
      "description": "The ranking column to use for selection.",
      "x-eval": "dict(enum=[f'{x}: {node.algorithms[\"ordering\"][x][\"algorithm_name\"]}' for x in list(node.ordering_result['rank_keys'])], default=[f'{x}: {node.algorithms[\"ordering\"][x][\"algorithm_name\"]}' for x in list(node.ordering_result['rank_keys'])][0])"
    },
    "comparison_operator": {
      "type": "string",
      "enum": [
        "==",
        "<=",
        ">=",
        "<",
        ">"
      ],
      "description": "The comparison operator to use for the ranking scores.",
      "default": "=="
    },
    "value": {
      "type": "number",
      "description": "The numeric value to compare the ranking scores against.",
      "default": 0
    }
  },
  "required": []
}

---

Select by Token-Level Numeric Attribute

Path: flexiconc/algorithms/select_by_token_numeric_value.py

Description:

Selects lines based on a token-level attribute using numeric comparison at a given offset. If a list is provided for 'value', only equality comparison is performed.

Arguments:

Name	Type	Description
value	['number', 'array']	The numeric value(s) to compare against. If a list is provided, only equality comparison is supported.
tokens_attribute	string	The token-level attribute to check.
offset	integer	The token offset to check.
comparison_operator	string	The comparison operator to use for numeric values. Ignored if 'value' is a list.
negative	boolean	If True, inverts the selection.

Show full JSON schema

{
  "type": "object",
  "properties": {
    "value": {
      "type": [
        "number",
        "array"
      ],
      "items": {
        "type": "number"
      },
      "description": "The numeric value(s) to compare against. If a list is provided, only equality comparison is supported.",
      "default": 0
    },
    "tokens_attribute": {
      "type": "string",
      "description": "The token-level attribute to check.",
      "x-eval": "dict(enum=[col for col in list(conc.tokens.columns) if col not in {'id_in_line', 'line_id', 'offset'} and ('int' in str(conc.tokens[col].dtype) or 'float' in str(conc.tokens[col].dtype))])"
    },
    "offset": {
      "type": "integer",
      "description": "The token offset to check.",
      "default": 0,
      "x-eval": "dict(minimum=min(conc.tokens['offset']), maximum=max(conc.tokens['offset']))"
    },
    "comparison_operator": {
      "type": "string",
      "enum": [
        "==",
        "<",
        ">",
        "<=",
        ">="
      ],
      "description": "The comparison operator to use for numeric values. Ignored if 'value' is a list.",
      "default": "=="
    },
    "negative": {
      "type": "boolean",
      "description": "If True, inverts the selection.",
      "default": false
    }
  },
  "required": [
    "value",
    "tokens_attribute"
  ]
}

---

Select by Token-Level String Attribute

Path: flexiconc/algorithms/select_by_token_string.py

Description:

Selects lines based on a token-level attribute (string matching) at a given offset. Supports regex and case sensitivity. The search_terms argument is a list of strings to match against.

Arguments:

Name	Type	Description
search_terms	array	The list of string values to match against.
tokens_attribute	string	The token attribute to check (e.g., 'word').
offset	integer	The token offset to check.
case_sensitive	boolean	If True, performs a case-sensitive match.
regex	boolean	If True, uses regex matching.
negative	boolean	If True, inverts the selection.

Show full JSON schema

{
  "type": "object",
  "properties": {
    "search_terms": {
      "type": "array",
      "items": {
        "type": "string"
      },
      "description": "The list of string values to match against.",
      "default": []
    },
    "tokens_attribute": {
      "type": "string",
      "description": "The token attribute to check (e.g., 'word').",
      "default": "word",
      "x-eval": "dict(enum=list(set(conc.tokens.columns) - {'id_in_line', 'line_id', 'offset'}))"
    },
    "offset": {
      "type": "integer",
      "description": "The token offset to check.",
      "default": 0,
      "x-eval": "dict(minimum=min(conc.tokens['offset']), maximum=max(conc.tokens['offset']))"
    },
    "case_sensitive": {
      "type": "boolean",
      "description": "If True, performs a case-sensitive match.",
      "default": false
    },
    "regex": {
      "type": "boolean",
      "description": "If True, uses regex matching.",
      "default": false
    },
    "negative": {
      "type": "boolean",
      "description": "If True, inverts the selection.",
      "default": false
    }
  },
  "required": [
    "search_terms"
  ]
}

---

Manual Line Selection

Path: flexiconc/algorithms/select_manual.py

Description:

Manually selects lines into a subset by specifying line IDs or groups (partitions or clusters) from the active node's grouping result. Additionally, ensures selection is restricted to allowed lines.

Arguments:

Name	Type	Description
line_ids	array	A list of specific line IDs to include in the subset.
groups	array	A list of group identifiers (by label or number) to include lines from. For clusters, groups may be nested, and all matching groups in the hierarchy will be used.

Show full JSON schema

{
  "type": "object",
  "properties": {
    "line_ids": {
      "type": "array",
      "items": {
        "type": "integer"
      },
      "description": "A list of specific line IDs to include in the subset."
    },
    "groups": {
      "type": "array",
      "items": {
        "type": [
          "string",
          "integer"
        ]
      },
      "description": "A list of group identifiers (by label or number) to include lines from. For clusters, groups may be nested, and all matching groups in the hierarchy will be used."
    }
  },
  "required": []
}

---

Random Sample

Path: flexiconc/algorithms/select_random.py

Description:

Selects a random sample of lines from the concordance, optionally using a seed.

Arguments:

Name	Type	Description
sample_size	integer	The number of lines to sample.
seed	integer	The seed for random number generation.

Show full JSON schema

{
  "type": "object",
  "properties": {
    "sample_size": {
      "type": "integer",
      "description": "The number of lines to sample.",
      "minimum": 1,
      "x-eval": "dict(maximum=node.line_count)"
    },
    "seed": {
      "type": "integer",
      "description": "The seed for random number generation.",
      "default": 42
    }
  },
  "required": [
    "sample_size"
  ]
}

---

Select Slot

Path: flexiconc/algorithms/select_slot.py

Description:

Selects the slot to work with.

Arguments:

Name	Type	Description
slot_id	integer	The slot identifier to select.

Show full JSON schema

{
  "type": "object",
  "properties": {
    "slot_id": {
      "type": "integer",
      "description": "The slot identifier to select.",
      "x-eval": "dict(enum=list(set(conc.matches['slot'])))"
    }
  },
  "required": [
    "slot_id"
  ]
}

---

Select Weighted Sample by Metadata

Path: flexiconc/algorithms/select_weighted_sample_by_metadata.py

Description:

Selects a weighted sample of lines based on the distribution of a specified metadata attribute.

Arguments:

Name	Type	Description
metadata_attribute	string	The metadata attribute to stratify by (e.g., 'text_id', 'speaker').
sample_size	integer	The total number of lines to sample.
seed	['integer']	An optional seed for generating the pseudo-random order.

Show full JSON schema

{
  "type": "object",
  "properties": {
    "metadata_attribute": {
      "type": "string",
      "description": "The metadata attribute to stratify by (e.g., 'text_id', 'speaker').",
      "x-eval": "dict(enum=list(set(conc.metadata.columns) - {'line_id'}))"
    },
    "sample_size": {
      "type": "integer",
      "description": "The total number of lines to sample.",
      "minimum": 1,
      "x-eval": "dict(maximum=node.line_count)"
    },
    "seed": {
      "type": [
        "integer"
      ],
      "description": "An optional seed for generating the pseudo-random order.",
      "default": 42
    }
  },
  "required": [
    "metadata_attribute",
    "sample_size"
  ]
}

---