DecisionMatrix Decorator¶

Structures the response as a decision matrix, evaluating options against multiple criteria. This decorator facilitates systematic comparison and selection between alternatives based on weighted or unweighted criteria.

Category: Structure

Parameters¶

Parameter	Type	Description	Default
`options`	`array`	Specific options or alternatives to evaluate in the matrix	``
`criteria`	`array`	Evaluation criteria to assess each option against	``
`weighted`	`boolean`	Whether to include weights for criteria importance	`False`
`scale`	`enum`	Rating scale to use for evaluations	`1-5`

Scale Options¶

1-5: Use a 1-5 rating scale for each criterion (where 1 is poor/lowest and 5 is excellent/highest).
1-10: Use a 1-10 rating scale for each criterion (where 1 is poor/lowest and 10 is excellent/highest).
qualitative: Use qualitative ratings (Poor, Fair, Good, Very Good, Excellent) for each criterion.
percentage: Use percentage scores (0-100%) for rating each criterion.

Examples¶

Simple decision matrix for comparing options¶

+++DecisionMatrix
What smartphone should I buy?

Creates a decision matrix comparing top smartphone options against key purchasing criteria, with 1-5 ratings for each combination

Weighted decision matrix with custom options and criteria¶

+++DecisionMatrix(options=[Python,JavaScript,Go,Rust], criteria=[learning curve,performance,ecosystem,job market], weighted=true, scale=1-10)
Which programming language should I learn next?

Generates a weighted decision matrix comparing the specified programming languages against the given criteria, with weighted scores on a 1-10 scale

Model-Specific Implementations¶

gpt-4-turbo¶

Instruction: Create a decision matrix comparing {options} against {criteria}. Rate each option-criterion pair using a {scale} scale. {weighted} Include a brief explanation for each rating, and conclude with a recommendation based on the matrix results.

Notes: This model sometimes needs more explicit instructions about formatting the matrix clearly and providing final scores

Implementation Guidance¶

Simple smartphone comparison¶

Original Prompt:

What smartphone should I buy?

Transformed Prompt:

Please structure your response as a decision matrix that systematically evaluates options against multiple criteria to facilitate comparison and selection. Use a 1-5 rating scale for each criterion (where 1 is poor/lowest and 5 is excellent/highest). Evaluate all criteria with equal importance, without applying weights to the scores.

What smartphone should I buy?

Detailed programming language comparison with weights¶

Original Prompt:

Which programming language should I learn next?

Transformed Prompt:

Please structure your response as a decision matrix that systematically evaluates options against multiple criteria to facilitate comparison and selection. Evaluate these specific options or alternatives in your matrix: Python, JavaScript, Go, Rust. Assess each option against these specific criteria: learning curve, performance, ecosystem, job market. Include weight factors for each criterion to reflect their relative importance, and calculate weighted scores for each option. Use a 1-10 rating scale for each criterion (where 1 is poor/lowest and 10 is excellent/highest).

Which programming language should I learn next?

Transformation Details¶

Base Instruction: Please structure your response as a decision matrix that systematically evaluates options against multiple criteria to facilitate comparison and selection.

Placement: prepend

Composition Behavior: accumulate

Parameter Effects:

options:
Format: Evaluate these specific options or alternatives in your matrix: {value}.
criteria:
Format: Assess each option against these specific criteria: {value}.
weighted:
When set to true: Include weight factors for each criterion to reflect their relative importance, and calculate weighted scores for each option.
When set to false: Evaluate all criteria with equal importance, without applying weights to the scores.
scale:
When set to 1-5: Use a 1-5 rating scale for each criterion (where 1 is poor/lowest and 5 is excellent/highest).
When set to 1-10: Use a 1-10 rating scale for each criterion (where 1 is poor/lowest and 10 is excellent/highest).
When set to qualitative: Use qualitative ratings (Poor, Fair, Good, Very Good, Excellent) for each criterion.
When set to percentage: Use percentage scores (0-100%) for rating each criterion.

Compatibility¶

Requires: None
Conflicts: None
Compatible Models: gpt-4o, gpt-4-turbo
Standard Version: 1.0.0 - 2.0.0

TableFormat: Enhances DecisionMatrix TableFormat can provide additional formatting options for the decision matrix presentation
Comparison: Enhances DecisionMatrix Comparison works well with DecisionMatrix by providing structure for qualitative comparisons alongside the quantitative matrix
OutputFormat: Enhances DecisionMatrix OutputFormat can be used to specify the format for the matrix (e.g., as markdown or CSV)