Implement MagnitudeVector and add corresponding test cases

[SYaoJun]

What changes were proposed in this pull request?

fix: #773

add MagnitudeVector with match logic and unit tests

How was this PR tested?

• Tests have Added for the changes
• Production environment verified

[SYaoJun]

@Leomrlin Could you please take a look at this code when you have a moment?

[kitalkuyo-gita]

Code snippet: if (!(other instanceof MagnitudeVector)) { throw new IllegalArgumentException("Other vector must be a MagnitudeVector"); }

Problem: This can cause runtime exceptions when calling match without explicitly performing type checking when combining multiple IVectors (EmbeddingVector, KeywordVector, TraversalVector, etc.) in VectorSearch, potentially interrupting the search process.

Recommendation: For IVectors of different types, it's advisable to return 0.0 (indicating dissimilarity/not included in the scoring), or have a more explicit contract defined by the IVector interface (e.g., match supports different types and returns 0). Returning 0 is more robust and consistent with the design comment in paste ("if not same type return 0.0").

[kitalkuyo-gita]

In the code, "magnitude" typically refers to metrics such as degree or PageRank—values ​​that are generally non-negative (≥0)—though certain centrality measures (e.g., specific eigenvector-based or normalized metrics) can be negative. The current implementation handles this using Math.abs, ensuring the calculation proceeds; however, it is advisable to either explicitly define the valid range (≥0) in the class documentation or formally document the meaning of negative values ​​if they are permitted.

Recommendation: Add Javadoc to the constructor specifying the expected range, or implement validation within the constructor (throwing an IllegalArgumentException if only non-negative values ​​are allowed).

[SYaoJun]

Good suggestions. I have updated the code. Please take a look again!

[SYaoJun]

What changes were proposed in this pull request?

fix: #773

add MagnitudeVector with match logic and unit tests

How was this PR tested?

• Tests have Added for the changes
• Production environment verified

[SYaoJun]

@Leomrlin Could you please take a look at this code when you have a moment?

[kitalkuyo-gita]

Code snippet: if (!(other instanceof MagnitudeVector)) { throw new IllegalArgumentException("Other vector must be a MagnitudeVector"); }

Problem: This can cause runtime exceptions when calling match without explicitly performing type checking when combining multiple IVectors (EmbeddingVector, KeywordVector, TraversalVector, etc.) in VectorSearch, potentially interrupting the search process.

Recommendation: For IVectors of different types, it's advisable to return 0.0 (indicating dissimilarity/not included in the scoring), or have a more explicit contract defined by the IVector interface (e.g., match supports different types and returns 0). Returning 0 is more robust and consistent with the design comment in paste ("if not same type return 0.0").

[kitalkuyo-gita]

In the code, "magnitude" typically refers to metrics such as degree or PageRank—values ​​that are generally non-negative (≥0)—though certain centrality measures (e.g., specific eigenvector-based or normalized metrics) can be negative. The current implementation handles this using Math.abs, ensuring the calculation proceeds; however, it is advisable to either explicitly define the valid range (≥0) in the class documentation or formally document the meaning of negative values ​​if they are permitted.

Recommendation: Add Javadoc to the constructor specifying the expected range, or implement validation within the constructor (throwing an IllegalArgumentException if only non-negative values ​​are allowed).

[SYaoJun]

Good suggestions. I have updated the code. Please take a look again!

[SYaoJun]

What changes were proposed in this pull request?

fix: #773

add MagnitudeVector with match logic and unit tests

How was this PR tested?

• Tests have Added for the changes
• Production environment verified

[SYaoJun]

@Leomrlin Could you please take a look at this code when you have a moment?

[kitalkuyo-gita]

Code snippet: if (!(other instanceof MagnitudeVector)) { throw new IllegalArgumentException("Other vector must be a MagnitudeVector"); }

Problem: This can cause runtime exceptions when calling match without explicitly performing type checking when combining multiple IVectors (EmbeddingVector, KeywordVector, TraversalVector, etc.) in VectorSearch, potentially interrupting the search process.

Recommendation: For IVectors of different types, it's advisable to return 0.0 (indicating dissimilarity/not included in the scoring), or have a more explicit contract defined by the IVector interface (e.g., match supports different types and returns 0). Returning 0 is more robust and consistent with the design comment in paste ("if not same type return 0.0").

[kitalkuyo-gita]

In the code, "magnitude" typically refers to metrics such as degree or PageRank—values ​​that are generally non-negative (≥0)—though certain centrality measures (e.g., specific eigenvector-based or normalized metrics) can be negative. The current implementation handles this using Math.abs, ensuring the calculation proceeds; however, it is advisable to either explicitly define the valid range (≥0) in the class documentation or formally document the meaning of negative values ​​if they are permitted.

Recommendation: Add Javadoc to the constructor specifying the expected range, or implement validation within the constructor (throwing an IllegalArgumentException if only non-negative values ​​are allowed).

[SYaoJun]

Good suggestions. I have updated the code. Please take a look again!

[SYaoJun]

What changes were proposed in this pull request?

fix: #773

add MagnitudeVector with match logic and unit tests

How was this PR tested?

• Tests have Added for the changes
• Production environment verified

[SYaoJun]

@Leomrlin Could you please take a look at this code when you have a moment?

[kitalkuyo-gita]

Code snippet: if (!(other instanceof MagnitudeVector)) { throw new IllegalArgumentException("Other vector must be a MagnitudeVector"); }

Problem: This can cause runtime exceptions when calling match without explicitly performing type checking when combining multiple IVectors (EmbeddingVector, KeywordVector, TraversalVector, etc.) in VectorSearch, potentially interrupting the search process.

Recommendation: For IVectors of different types, it's advisable to return 0.0 (indicating dissimilarity/not included in the scoring), or have a more explicit contract defined by the IVector interface (e.g., match supports different types and returns 0). Returning 0 is more robust and consistent with the design comment in paste ("if not same type return 0.0").

[kitalkuyo-gita]

In the code, "magnitude" typically refers to metrics such as degree or PageRank—values ​​that are generally non-negative (≥0)—though certain centrality measures (e.g., specific eigenvector-based or normalized metrics) can be negative. The current implementation handles this using Math.abs, ensuring the calculation proceeds; however, it is advisable to either explicitly define the valid range (≥0) in the class documentation or formally document the meaning of negative values ​​if they are permitted.

Recommendation: Add Javadoc to the constructor specifying the expected range, or implement validation within the constructor (throwing an IllegalArgumentException if only non-negative values ​​are allowed).

[SYaoJun]

Good suggestions. I have updated the code. Please take a look again!

[SYaoJun]

What changes were proposed in this pull request?

fix: #773

add MagnitudeVector with match logic and unit tests

How was this PR tested?

• Tests have Added for the changes
• Production environment verified

[SYaoJun]

@Leomrlin Could you please take a look at this code when you have a moment?

[kitalkuyo-gita]

Code snippet: if (!(other instanceof MagnitudeVector)) { throw new IllegalArgumentException("Other vector must be a MagnitudeVector"); }

Problem: This can cause runtime exceptions when calling match without explicitly performing type checking when combining multiple IVectors (EmbeddingVector, KeywordVector, TraversalVector, etc.) in VectorSearch, potentially interrupting the search process.

Recommendation: For IVectors of different types, it's advisable to return 0.0 (indicating dissimilarity/not included in the scoring), or have a more explicit contract defined by the IVector interface (e.g., match supports different types and returns 0). Returning 0 is more robust and consistent with the design comment in paste ("if not same type return 0.0").

[kitalkuyo-gita]

In the code, "magnitude" typically refers to metrics such as degree or PageRank—values ​​that are generally non-negative (≥0)—though certain centrality measures (e.g., specific eigenvector-based or normalized metrics) can be negative. The current implementation handles this using Math.abs, ensuring the calculation proceeds; however, it is advisable to either explicitly define the valid range (≥0) in the class documentation or formally document the meaning of negative values ​​if they are permitted.

Recommendation: Add Javadoc to the constructor specifying the expected range, or implement validation within the constructor (throwing an IllegalArgumentException if only non-negative values ​​are allowed).

[SYaoJun]

Good suggestions. I have updated the code. Please take a look again!

[SYaoJun]

What changes were proposed in this pull request?

fix: #773

add MagnitudeVector with match logic and unit tests

How was this PR tested?

• Tests have Added for the changes
• Production environment verified

[SYaoJun]

@Leomrlin Could you please take a look at this code when you have a moment?

[kitalkuyo-gita]

Code snippet: if (!(other instanceof MagnitudeVector)) { throw new IllegalArgumentException("Other vector must be a MagnitudeVector"); }

Problem: This can cause runtime exceptions when calling match without explicitly performing type checking when combining multiple IVectors (EmbeddingVector, KeywordVector, TraversalVector, etc.) in VectorSearch, potentially interrupting the search process.

Recommendation: For IVectors of different types, it's advisable to return 0.0 (indicating dissimilarity/not included in the scoring), or have a more explicit contract defined by the IVector interface (e.g., match supports different types and returns 0). Returning 0 is more robust and consistent with the design comment in paste ("if not same type return 0.0").

[kitalkuyo-gita]

In the code, "magnitude" typically refers to metrics such as degree or PageRank—values ​​that are generally non-negative (≥0)—though certain centrality measures (e.g., specific eigenvector-based or normalized metrics) can be negative. The current implementation handles this using Math.abs, ensuring the calculation proceeds; however, it is advisable to either explicitly define the valid range (≥0) in the class documentation or formally document the meaning of negative values ​​if they are permitted.

Recommendation: Add Javadoc to the constructor specifying the expected range, or implement validation within the constructor (throwing an IllegalArgumentException if only non-negative values ​​are allowed).

[SYaoJun]

Good suggestions. I have updated the code. Please take a look again!

[SYaoJun]

What changes were proposed in this pull request?

fix: #773

add MagnitudeVector with match logic and unit tests

How was this PR tested?

• Tests have Added for the changes
• Production environment verified

[SYaoJun]

@Leomrlin Could you please take a look at this code when you have a moment?

[kitalkuyo-gita]

Code snippet: if (!(other instanceof MagnitudeVector)) { throw new IllegalArgumentException("Other vector must be a MagnitudeVector"); }

Problem: This can cause runtime exceptions when calling match without explicitly performing type checking when combining multiple IVectors (EmbeddingVector, KeywordVector, TraversalVector, etc.) in VectorSearch, potentially interrupting the search process.

Recommendation: For IVectors of different types, it's advisable to return 0.0 (indicating dissimilarity/not included in the scoring), or have a more explicit contract defined by the IVector interface (e.g., match supports different types and returns 0). Returning 0 is more robust and consistent with the design comment in paste ("if not same type return 0.0").

[kitalkuyo-gita]

In the code, "magnitude" typically refers to metrics such as degree or PageRank—values ​​that are generally non-negative (≥0)—though certain centrality measures (e.g., specific eigenvector-based or normalized metrics) can be negative. The current implementation handles this using Math.abs, ensuring the calculation proceeds; however, it is advisable to either explicitly define the valid range (≥0) in the class documentation or formally document the meaning of negative values ​​if they are permitted.

Recommendation: Add Javadoc to the constructor specifying the expected range, or implement validation within the constructor (throwing an IllegalArgumentException if only non-negative values ​​are allowed).

[SYaoJun]

Good suggestions. I have updated the code. Please take a look again!

[SYaoJun]

What changes were proposed in this pull request?

fix: #773

add MagnitudeVector with match logic and unit tests

How was this PR tested?

• Tests have Added for the changes
• Production environment verified

[SYaoJun]

@Leomrlin Could you please take a look at this code when you have a moment?

[kitalkuyo-gita]

Code snippet: if (!(other instanceof MagnitudeVector)) { throw new IllegalArgumentException("Other vector must be a MagnitudeVector"); }

Problem: This can cause runtime exceptions when calling match without explicitly performing type checking when combining multiple IVectors (EmbeddingVector, KeywordVector, TraversalVector, etc.) in VectorSearch, potentially interrupting the search process.

Recommendation: For IVectors of different types, it's advisable to return 0.0 (indicating dissimilarity/not included in the scoring), or have a more explicit contract defined by the IVector interface (e.g., match supports different types and returns 0). Returning 0 is more robust and consistent with the design comment in paste ("if not same type return 0.0").

[kitalkuyo-gita]

In the code, "magnitude" typically refers to metrics such as degree or PageRank—values ​​that are generally non-negative (≥0)—though certain centrality measures (e.g., specific eigenvector-based or normalized metrics) can be negative. The current implementation handles this using Math.abs, ensuring the calculation proceeds; however, it is advisable to either explicitly define the valid range (≥0) in the class documentation or formally document the meaning of negative values ​​if they are permitted.

Recommendation: Add Javadoc to the constructor specifying the expected range, or implement validation within the constructor (throwing an IllegalArgumentException if only non-negative values ​​are allowed).

[SYaoJun]

Good suggestions. I have updated the code. Please take a look again!