Add logSumExp and logSumExpPair functions

[ohikouta]

Summary

Implements numerically stable log-sum-exp computation, as requested in #65.

• logSumExp :: U.Vector Double -> Double — computes log(sum(exp(x_i))) using the max-subtraction trick to avoid overflow/underflow
• logSumExpPair :: Double -> Double -> Double — optimized two-argument version using log1p for higher precision

These functions are commonly used when working with log-probabilities in statistical computing and machine learning.

Changes

• Numeric.SpecFunctions.Internal: implementation of both functions
• Numeric.SpecFunctions: exported under a new "Log-sum-exp" section
• tests/Tests/SpecFunctions.hs: property tests (singleton, commutativity, identity) and unit tests (empty vector, large values, log-probability combination)

Test plan

• All 74 tests pass (cabal test)
• Existing tests unaffected
• Property-based tests verify numerical correctness
• Edge cases: empty vector returns -Infinity, large values (1000+) handled correctly

Closes#65

[ohikouta]

Summary

Implements numerically stable log-sum-exp computation, as requested in #65.

• logSumExp :: U.Vector Double -> Double — computes log(sum(exp(x_i))) using the max-subtraction trick to avoid overflow/underflow
• logSumExpPair :: Double -> Double -> Double — optimized two-argument version using log1p for higher precision

These functions are commonly used when working with log-probabilities in statistical computing and machine learning.

Changes

• Numeric.SpecFunctions.Internal: implementation of both functions
• Numeric.SpecFunctions: exported under a new "Log-sum-exp" section
• tests/Tests/SpecFunctions.hs: property tests (singleton, commutativity, identity) and unit tests (empty vector, large values, log-probability combination)

Test plan

• All 74 tests pass (cabal test)
• Existing tests unaffected
• Property-based tests verify numerical correctness
• Edge cases: empty vector returns -Infinity, large values (1000+) handled correctly

Closes#65

[ohikouta]

Summary

Implements numerically stable log-sum-exp computation, as requested in #65.

• logSumExp :: U.Vector Double -> Double — computes log(sum(exp(x_i))) using the max-subtraction trick to avoid overflow/underflow
• logSumExpPair :: Double -> Double -> Double — optimized two-argument version using log1p for higher precision

These functions are commonly used when working with log-probabilities in statistical computing and machine learning.

Changes

• Numeric.SpecFunctions.Internal: implementation of both functions
• Numeric.SpecFunctions: exported under a new "Log-sum-exp" section
• tests/Tests/SpecFunctions.hs: property tests (singleton, commutativity, identity) and unit tests (empty vector, large values, log-probability combination)

Test plan

• All 74 tests pass (cabal test)
• Existing tests unaffected
• Property-based tests verify numerical correctness
• Edge cases: empty vector returns -Infinity, large values (1000+) handled correctly

Closes#65

[ohikouta]

Summary

Implements numerically stable log-sum-exp computation, as requested in #65.

• logSumExp :: U.Vector Double -> Double — computes log(sum(exp(x_i))) using the max-subtraction trick to avoid overflow/underflow
• logSumExpPair :: Double -> Double -> Double — optimized two-argument version using log1p for higher precision

These functions are commonly used when working with log-probabilities in statistical computing and machine learning.

Changes

• Numeric.SpecFunctions.Internal: implementation of both functions
• Numeric.SpecFunctions: exported under a new "Log-sum-exp" section
• tests/Tests/SpecFunctions.hs: property tests (singleton, commutativity, identity) and unit tests (empty vector, large values, log-probability combination)

Test plan

• All 74 tests pass (cabal test)
• Existing tests unaffected
• Property-based tests verify numerical correctness
• Edge cases: empty vector returns -Infinity, large values (1000+) handled correctly

Closes#65

[ohikouta]

Summary

Implements numerically stable log-sum-exp computation, as requested in #65.

• logSumExp :: U.Vector Double -> Double — computes log(sum(exp(x_i))) using the max-subtraction trick to avoid overflow/underflow
• logSumExpPair :: Double -> Double -> Double — optimized two-argument version using log1p for higher precision

These functions are commonly used when working with log-probabilities in statistical computing and machine learning.

Changes

• Numeric.SpecFunctions.Internal: implementation of both functions
• Numeric.SpecFunctions: exported under a new "Log-sum-exp" section
• tests/Tests/SpecFunctions.hs: property tests (singleton, commutativity, identity) and unit tests (empty vector, large values, log-probability combination)

Test plan

• All 74 tests pass (cabal test)
• Existing tests unaffected
• Property-based tests verify numerical correctness
• Edge cases: empty vector returns -Infinity, large values (1000+) handled correctly

Closes#65

[ohikouta]

Summary

Implements numerically stable log-sum-exp computation, as requested in #65.

• logSumExp :: U.Vector Double -> Double — computes log(sum(exp(x_i))) using the max-subtraction trick to avoid overflow/underflow
• logSumExpPair :: Double -> Double -> Double — optimized two-argument version using log1p for higher precision

These functions are commonly used when working with log-probabilities in statistical computing and machine learning.

Changes

• Numeric.SpecFunctions.Internal: implementation of both functions
• Numeric.SpecFunctions: exported under a new "Log-sum-exp" section
• tests/Tests/SpecFunctions.hs: property tests (singleton, commutativity, identity) and unit tests (empty vector, large values, log-probability combination)

Test plan

• All 74 tests pass (cabal test)
• Existing tests unaffected
• Property-based tests verify numerical correctness
• Edge cases: empty vector returns -Infinity, large values (1000+) handled correctly

Closes#65

[ohikouta]

Summary

Implements numerically stable log-sum-exp computation, as requested in #65.

• logSumExp :: U.Vector Double -> Double — computes log(sum(exp(x_i))) using the max-subtraction trick to avoid overflow/underflow
• logSumExpPair :: Double -> Double -> Double — optimized two-argument version using log1p for higher precision

These functions are commonly used when working with log-probabilities in statistical computing and machine learning.

Changes

• Numeric.SpecFunctions.Internal: implementation of both functions
• Numeric.SpecFunctions: exported under a new "Log-sum-exp" section
• tests/Tests/SpecFunctions.hs: property tests (singleton, commutativity, identity) and unit tests (empty vector, large values, log-probability combination)

Test plan

• All 74 tests pass (cabal test)
• Existing tests unaffected
• Property-based tests verify numerical correctness
• Edge cases: empty vector returns -Infinity, large values (1000+) handled correctly

Closes#65

[ohikouta]

Summary

Implements numerically stable log-sum-exp computation, as requested in #65.

• logSumExp :: U.Vector Double -> Double — computes log(sum(exp(x_i))) using the max-subtraction trick to avoid overflow/underflow
• logSumExpPair :: Double -> Double -> Double — optimized two-argument version using log1p for higher precision

These functions are commonly used when working with log-probabilities in statistical computing and machine learning.

Changes

• Numeric.SpecFunctions.Internal: implementation of both functions
• Numeric.SpecFunctions: exported under a new "Log-sum-exp" section
• tests/Tests/SpecFunctions.hs: property tests (singleton, commutativity, identity) and unit tests (empty vector, large values, log-probability combination)

Test plan

• All 74 tests pass (cabal test)
• Existing tests unaffected
• Property-based tests verify numerical correctness
• Edge cases: empty vector returns -Infinity, large values (1000+) handled correctly

Closes#65