diff --git a/lib/node_modules/@stdlib/blas/ext/base/dasumpw/README.md b/lib/node_modules/@stdlib/blas/ext/base/dasumpw/README.md index 0409e7d63c78..f53c75b157c6 100644 --- a/lib/node_modules/@stdlib/blas/ext/base/dasumpw/README.md +++ b/lib/node_modules/@stdlib/blas/ext/base/dasumpw/README.md @@ -51,7 +51,7 @@ The [_L1_ norm][l1norm] is defined as var dasumpw = require( '@stdlib/blas/ext/base/dasumpw' ); ``` -#### dasumpw( N, x, stride ) +#### dasumpw( N, x, strideX ) Computes the sum of absolute values ([_L1_ norm][l1norm]) of double-precision floating-point strided array elements using pairwise summation. @@ -69,7 +69,7 @@ The function has the following parameters: - **N**: number of indexed elements. - **x**: input [`Float64Array`][@stdlib/array/float64]. -- **stride**: index increment for `x`. +- **strideX**: index increment for `x`. The `N` and stride parameters determine which elements in the strided array are accessed at runtime. For example, to compute the sum of absolute values of every other element in `x`, @@ -96,7 +96,7 @@ var v = dasumpw( 4, x1, 2 ); // returns 9.0 ``` -#### dasumpw.ndarray( N, x, stride, offset ) +#### dasumpw.ndarray( N, x, strideX, offsetX ) Computes the sum of absolute values ([_L1_ norm][l1norm]) of double-precision floating-point strided array elements using pairwise summation and alternative indexing semantics. @@ -112,9 +112,9 @@ var v = dasumpw.ndarray( N, x, 1, 0 ); The function has the following additional parameters: -- **offset**: starting index for `x`. +- **offsetX**: starting index for `x`. -While [`typed array`][mdn-typed-array] views mandate a view offset based on the underlying `buffer`, the `offset` parameter supports indexing semantics based on a starting index. For example, to calculate the sum of absolute values of every other value in `x` starting from the second value +While [`typed array`][mdn-typed-array] views mandate a view offset based on the underlying buffer, the offset parameter supports indexing semantics based on a starting index. For example, to calculate the sum of absolute values of every other value in `x` starting from the second value ```javascript var Float64Array = require( '@stdlib/array/float64' ); @@ -147,11 +147,12 @@ var v = dasumpw.ndarray( 4, x, 2, 1 ); ```javascript -var discreteUniform = require( '@stdlib/random/base/discrete-uniform' ).factory; -var filledarrayBy = require( '@stdlib/array/filled-by' ); +var discreteUniform = require( '@stdlib/random/array/discrete-uniform' ); var dasumpw = require( '@stdlib/blas/ext/base/dasumpw' ); -var x = filledarrayBy( 10, 'float64', discreteUniform( 0, 100 ) ); +var x = discreteUniform( 10, -100, 100, { + 'dtype': 'float64' +}); console.log( x ); var v = dasumpw( x.length, x, 1 ); @@ -162,8 +163,123 @@ console.log( v ); + + * * * +
+ +## C APIs + + + +
+ +
+ + + + + +
+ +### Usage + +```c +#include "stdlib/blas/ext/base/dasumpw.h" +``` + +#### stdlib_strided_dasumpw( N, \*X, strideX ) + +Computes the sum of absolute values ([_L1_ norm][l1norm]) of double-precision floating-point strided array elements using pairwise summation. + +```c +const double x[] = { 1.0, 2.0, 3.0, 4.0 } + +double v = stdlib_strided_dasumpw( 4, x, 1 ); +// returns 10.0 +``` + +The function accepts the following arguments: + +- **N**: `[in] CBLAS_INT` number of indexed elements. +- **X**: `[in] double*` input array. +- **strideX**: `[in] CBLAS_INT` index increment for `X`. + +```c +double stdlib_strided_dasumpw( const CBLAS_INT N, const double *X, const CBLAS_INT strideX ); +``` + +#### stdlib_strided_dasumpw_ndarray( N, \*X, strideX, offsetX ) + +Computes the sum of absolute values ([_L1_ norm][l1norm]) of double-precision floating-point strided array elements using pairwise summation and alternative indexing semantics. + +```c +const double x[] = { 1.0, 2.0, 3.0, 4.0 } + +double v = stdlib_strided_dasumpw_ndarray( 4, x, 1, 0 ); +// returns 10.0 +``` + +The function accepts the following arguments: + +- **N**: `[in] CBLAS_INT` number of indexed elements. +- **X**: `[in] double*` input array. +- **strideX**: `[in] CBLAS_INT` index increment for `X`. +- **offsetX**: `[in] CBLAS_INT` starting index for `X`. + +```c +double stdlib_strided_dasumpw_ndarray( const CBLAS_INT N, const double *X, const CBLAS_INT strideX, const CBLAS_INT offsetX ); +``` + +
+ + + + + +
+ +
+ + + + + +
+ +### Examples + +```c +#include "stdlib/blas/ext/base/dasumpw.h" +#include + +int main( void ) { + // Create a strided array: + const double x[] = { 1.0, -2.0, 3.0, -4.0, 5.0, -6.0, 7.0, -8.0 }; + + // Specify the number of indexed elements: + const int N = 8; + + // Specify a stride: + const int strideX = 1; + + // Compute the sum: + double v = stdlib_strided_dasumpw( N, x, strideX ); + + // Print the result: + printf( "sumabs: %lf\n", sum ); +} +``` + +
+ + + +
+ + +
## References diff --git a/lib/node_modules/@stdlib/blas/ext/base/dasumpw/benchmark/benchmark.js b/lib/node_modules/@stdlib/blas/ext/base/dasumpw/benchmark/benchmark.js index 90ab3e8a33bb..2cf39e4c40de 100644 --- a/lib/node_modules/@stdlib/blas/ext/base/dasumpw/benchmark/benchmark.js +++ b/lib/node_modules/@stdlib/blas/ext/base/dasumpw/benchmark/benchmark.js @@ -21,8 +21,7 @@ // MODULES // var bench = require( '@stdlib/bench' ); -var uniform = require( '@stdlib/random/base/uniform' ).factory; -var filledarrayBy = require( '@stdlib/array/filled-by' ); +var uniform = require( '@stdlib/random/array/uniform' ); var isnan = require( '@stdlib/math/base/assert/is-nan' ); var pow = require( '@stdlib/math/base/special/pow' ); var pkg = require( './../package.json' ).name; @@ -31,7 +30,9 @@ var dasumpw = require( './../lib/dasumpw.js' ); // VARIABLES // -var rand = uniform( -100.0, 100.0 ); +var options = { + 'dtype': 'float64' +}; // FUNCTIONS // @@ -44,7 +45,7 @@ var rand = uniform( -100.0, 100.0 ); * @returns {Function} benchmark function */ function createBenchmark( len ) { - var x = filledarrayBy( len, 'float64', rand ); + var x = uniform( len, -100, 100, options ); return benchmark; function benchmark( b ) { diff --git a/lib/node_modules/@stdlib/blas/ext/base/dasumpw/benchmark/benchmark.native.js b/lib/node_modules/@stdlib/blas/ext/base/dasumpw/benchmark/benchmark.native.js index 610dafc3a3e9..78b78ce79dfe 100644 --- a/lib/node_modules/@stdlib/blas/ext/base/dasumpw/benchmark/benchmark.native.js +++ b/lib/node_modules/@stdlib/blas/ext/base/dasumpw/benchmark/benchmark.native.js @@ -22,8 +22,7 @@ var resolve = require( 'path' ).resolve; var bench = require( '@stdlib/bench' ); -var uniform = require( '@stdlib/random/base/uniform' ).factory; -var filledarrayBy = require( '@stdlib/array/filled-by' ); +var uniform = require( '@stdlib/random/array/uniform' ); var isnan = require( '@stdlib/math/base/assert/is-nan' ); var pow = require( '@stdlib/math/base/special/pow' ); var tryRequire = require( '@stdlib/utils/try-require' ); @@ -36,7 +35,9 @@ var dasumpw = tryRequire( resolve( __dirname, './../lib/dasumpw.native.js' ) ); var opts = { 'skip': ( dasumpw instanceof Error ) }; -var rand = uniform( -100.0, 100.0 ); +var options = { + 'dtype': 'float64' +}; // FUNCTIONS // @@ -49,7 +50,7 @@ var rand = uniform( -100.0, 100.0 ); * @returns {Function} benchmark function */ function createBenchmark( len ) { - var x = filledarrayBy( len, 'float64', rand ); + var x = uniform( len, -100, 100, options ); return benchmark; function benchmark( b ) { diff --git a/lib/node_modules/@stdlib/blas/ext/base/dasumpw/benchmark/benchmark.ndarray.js b/lib/node_modules/@stdlib/blas/ext/base/dasumpw/benchmark/benchmark.ndarray.js index a171c49f3faa..89169ceeaa75 100644 --- a/lib/node_modules/@stdlib/blas/ext/base/dasumpw/benchmark/benchmark.ndarray.js +++ b/lib/node_modules/@stdlib/blas/ext/base/dasumpw/benchmark/benchmark.ndarray.js @@ -21,8 +21,7 @@ // MODULES // var bench = require( '@stdlib/bench' ); -var uniform = require( '@stdlib/random/base/uniform' ).factory; -var filledarrayBy = require( '@stdlib/array/filled-by' ); +var uniform = require( '@stdlib/random/array/uniform' ); var isnan = require( '@stdlib/math/base/assert/is-nan' ); var pow = require( '@stdlib/math/base/special/pow' ); var pkg = require( './../package.json' ).name; @@ -31,7 +30,9 @@ var dasumpw = require( './../lib/ndarray.js' ); // VARIABLES // -var rand = uniform( -100.0, 100.0 ); +var options = { + 'dtype': 'float64' +}; // FUNCTIONS // @@ -44,7 +45,7 @@ var rand = uniform( -100.0, 100.0 ); * @returns {Function} benchmark function */ function createBenchmark( len ) { - var x = filledarrayBy( len, 'float64', rand ); + var x = uniform( len, -100, 100, options ); return benchmark; function benchmark( b ) { diff --git a/lib/node_modules/@stdlib/blas/ext/base/dasumpw/benchmark/benchmark.ndarray.native.js b/lib/node_modules/@stdlib/blas/ext/base/dasumpw/benchmark/benchmark.ndarray.native.js index c9d57e4634d9..2a0b25985216 100644 --- a/lib/node_modules/@stdlib/blas/ext/base/dasumpw/benchmark/benchmark.ndarray.native.js +++ b/lib/node_modules/@stdlib/blas/ext/base/dasumpw/benchmark/benchmark.ndarray.native.js @@ -22,8 +22,7 @@ var resolve = require( 'path' ).resolve; var bench = require( '@stdlib/bench' ); -var uniform = require( '@stdlib/random/base/uniform' ).factory; -var filledarrayBy = require( '@stdlib/array/filled-by' ); +var uniform = require( '@stdlib/random/array/uniform' ); var isnan = require( '@stdlib/math/base/assert/is-nan' ); var pow = require( '@stdlib/math/base/special/pow' ); var tryRequire = require( '@stdlib/utils/try-require' ); @@ -36,7 +35,9 @@ var dasumpw = tryRequire( resolve( __dirname, './../lib/ndarray.native.js' ) ); var opts = { 'skip': ( dasumpw instanceof Error ) }; -var rand = uniform( -100.0, 100.0 ); +var options = { + 'dtype': 'float64' +}; // FUNCTIONS // @@ -49,7 +50,7 @@ var rand = uniform( -100.0, 100.0 ); * @returns {Function} benchmark function */ function createBenchmark( len ) { - var x = filledarrayBy( len, 'float64', rand ); + var x = uniform( len, -100, 100, options ); return benchmark; function benchmark( b ) { diff --git a/lib/node_modules/@stdlib/blas/ext/base/dasumpw/benchmark/c/benchmark.length.c b/lib/node_modules/@stdlib/blas/ext/base/dasumpw/benchmark/c/benchmark.length.c index c4b0aad45c1c..4e53354aa210 100644 --- a/lib/node_modules/@stdlib/blas/ext/base/dasumpw/benchmark/c/benchmark.length.c +++ b/lib/node_modules/@stdlib/blas/ext/base/dasumpw/benchmark/c/benchmark.length.c @@ -94,7 +94,7 @@ static double rand_double( void ) { * @param len array length * @return elapsed time in seconds */ -static double benchmark( int iterations, int len ) { +static double benchmark1( int iterations, int len ) { double elapsed; double x[ len ]; double v; @@ -120,6 +120,39 @@ static double benchmark( int iterations, int len ) { return elapsed; } +/** +* Runs a benchmark. +* +* @param iterations number of iterations +* @param len array length +* @return elapsed time in seconds +*/ +static double benchmark2( int iterations, int len ) { + double elapsed; + double x[ len ]; + double v; + double t; + int i; + + for ( i = 0; i < len; i++ ) { + x[ i ] = ( rand_double() * 20000.0 ) - 10000.0; + } + v = 0.0; + t = tic(); + for ( i = 0; i < iterations; i++ ) { + v = stdlib_strided_dasumpw_ndarray( len, x, 1, 0 ); + if ( v != v ) { + printf( "should not return NaN\n" ); + break; + } + } + elapsed = tic() - t; + if ( v != v ) { + printf( "should not return NaN\n" ); + } + return elapsed; +} + /** * Main execution sequence. */ @@ -142,7 +175,18 @@ int main( void ) { for ( j = 0; j < REPEATS; j++ ) { count += 1; printf( "# c::%s:len=%d\n", NAME, len ); - elapsed = benchmark( iter, len ); + elapsed = benchmark1( iter, len ); + print_results( iter, elapsed ); + printf( "ok %d benchmark finished\n", count ); + } + } + for ( i = MIN; i <= MAX; i++ ) { + len = pow( 10, i ); + iter = ITERATIONS / pow( 10, i-1 ); + for ( j = 0; j < REPEATS; j++ ) { + count += 1; + printf( "# c::%s:ndarray:len=%d\n", NAME, len ); + elapsed = benchmark2( iter, len ); print_results( iter, elapsed ); printf( "ok %d benchmark finished\n", count ); } diff --git a/lib/node_modules/@stdlib/blas/ext/base/dasumpw/docs/repl.txt b/lib/node_modules/@stdlib/blas/ext/base/dasumpw/docs/repl.txt index e9dac9703ee3..ec0f7c2d3622 100644 --- a/lib/node_modules/@stdlib/blas/ext/base/dasumpw/docs/repl.txt +++ b/lib/node_modules/@stdlib/blas/ext/base/dasumpw/docs/repl.txt @@ -1,9 +1,9 @@ -{{alias}}( N, x, stride ) +{{alias}}( N, x, strideX ) Computes the sum of absolute values (L1 norm) of double-precision floating- point strided array elements using pairwise summation. - The `N` and `stride` parameters determine which elements in `x` are accessed + The `N` and stride parameters determine which elements in `x` are accessed at runtime. Indexing is relative to the first index. To introduce an offset, use a typed @@ -19,7 +19,7 @@ x: Float64Array Input array. - stride: integer + strideX: integer Index increment. Returns @@ -49,13 +49,14 @@ > {{alias}}( N, x1, stride ) 5.0 -{{alias}}.ndarray( N, x, stride, offset ) + +{{alias}}.ndarray( N, x, strideX, offsetX ) Computes the sum of absolute values (L1 norm) of double-precision floating- point strided array elements using pairwise summation and alternative indexing semantics. While typed array views mandate a view offset based on the underlying - buffer, the `offset` parameter supports indexing semantics based on a + buffer, the offset parameter supports indexing semantics based on a starting index. Parameters @@ -66,10 +67,10 @@ x: Float64Array Input array. - stride: integer + strideX: integer Index increment. - offset: integer + offsetX: integer Starting index. Returns diff --git a/lib/node_modules/@stdlib/blas/ext/base/dasumpw/docs/types/index.d.ts b/lib/node_modules/@stdlib/blas/ext/base/dasumpw/docs/types/index.d.ts index 46aa92e688fc..e602c86e9aa3 100644 --- a/lib/node_modules/@stdlib/blas/ext/base/dasumpw/docs/types/index.d.ts +++ b/lib/node_modules/@stdlib/blas/ext/base/dasumpw/docs/types/index.d.ts @@ -27,7 +27,7 @@ interface Routine { * * @param N - number of indexed elements * @param x - input array - * @param stride - stride length + * @param strideX - stride length * @returns sum * * @example @@ -38,15 +38,15 @@ interface Routine { * var v = dasumpw( x.length, x, 1 ); * // returns 1.0 */ - ( N: number, x: Float64Array, stride: number ): number; + ( N: number, x: Float64Array, strideX: number ): number; /** * Computes the sum of absolute values (L1 norm) of double-precision floating-point strided array elements using pairwise summation and alternative indexing semantics. * * @param N - number of indexed elements * @param x - input array - * @param stride - stride length - * @param offset - starting index + * @param strideX - stride length + * @param offsetX - starting index * @returns sum * * @example @@ -57,7 +57,7 @@ interface Routine { * var v = dasumpw.ndarray( x.length, x, 1, 0 ); * // returns 1.0 */ - ndarray( N: number, x: Float64Array, stride: number, offset: number ): number; + ndarray( N: number, x: Float64Array, strideX: number, offsetX: number ): number; } /** @@ -65,7 +65,7 @@ interface Routine { * * @param N - number of indexed elements * @param x - input array -* @param stride - stride length +* @param strideX - stride length * @returns sum * * @example diff --git a/lib/node_modules/@stdlib/blas/ext/base/dasumpw/examples/c/example.c b/lib/node_modules/@stdlib/blas/ext/base/dasumpw/examples/c/example.c index efd596d56ef5..63aa4ea6792f 100644 --- a/lib/node_modules/@stdlib/blas/ext/base/dasumpw/examples/c/example.c +++ b/lib/node_modules/@stdlib/blas/ext/base/dasumpw/examples/c/example.c @@ -25,13 +25,13 @@ int main( void ) { const double x[] = { -1.0, 2.0, -3.0, 4.0, -5.0, 6.0, -7.0, 8.0 }; // Specify the number of elements: - const int64_t N = 4; + const int N = 4; // Specify the stride length: - const int64_t stride = 2; + const int strideX = 2; // Compute the sum: - double v = stdlib_strided_dasumpw( N, x, stride ); + double v = stdlib_strided_dasumpw( N, x, strideX ); // Print the result: printf( "sumabs: %lf\n", v ); diff --git a/lib/node_modules/@stdlib/blas/ext/base/dasumpw/examples/index.js b/lib/node_modules/@stdlib/blas/ext/base/dasumpw/examples/index.js index 96b5f7748224..e571e5ad7149 100644 --- a/lib/node_modules/@stdlib/blas/ext/base/dasumpw/examples/index.js +++ b/lib/node_modules/@stdlib/blas/ext/base/dasumpw/examples/index.js @@ -18,11 +18,12 @@ 'use strict'; -var discreteUniform = require( '@stdlib/random/base/discrete-uniform' ).factory; -var filledarrayBy = require( '@stdlib/array/filled-by' ); +var discreteUniform = require( '@stdlib/random/array/discrete-uniform' ); var dasumpw = require( './../lib' ); -var x = filledarrayBy( 10, 'float64', discreteUniform( -100.0, 100.0 ) ); +var x = discreteUniform( 10, -100, 100, { + 'dtype': 'float64' +}); console.log( x ); var v = dasumpw( x.length, x, 1 ); diff --git a/lib/node_modules/@stdlib/blas/ext/base/dasumpw/include/stdlib/blas/ext/base/dasumpw.h b/lib/node_modules/@stdlib/blas/ext/base/dasumpw/include/stdlib/blas/ext/base/dasumpw.h index 19ae380a684d..2c65395ce612 100644 --- a/lib/node_modules/@stdlib/blas/ext/base/dasumpw/include/stdlib/blas/ext/base/dasumpw.h +++ b/lib/node_modules/@stdlib/blas/ext/base/dasumpw/include/stdlib/blas/ext/base/dasumpw.h @@ -19,7 +19,7 @@ #ifndef STDLIB_BLAS_EXT_BASE_DASUMPW_H #define STDLIB_BLAS_EXT_BASE_DASUMPW_H -#include +#include "stdlib/blas/base/shared.h" /* * If C++, prevent name mangling so that the compiler emits a binary file having undecorated names, thus mirroring the behavior of a C compiler. @@ -31,7 +31,12 @@ extern "C" { /** * Computes the sum of absolute values (L1 norm) of double-precision floating-point strided array elements using pairwise summation. */ -double stdlib_strided_dasumpw( const int64_t N, const double *X, const int64_t stride ); +double API_SUFFIX(stdlib_strided_dasumpw)( const CBLAS_INT N, const double *X, const CBLAS_INT strideX ); + +/** +* Computes the sum of absolute values (L1 norm) of double-precision floating-point strided array elements using pairwise summation and alternative indexing semantics. +*/ +double API_SUFFIX(stdlib_strided_dasumpw_ndarray)( const CBLAS_INT N, const double *X, const CBLAS_INT strideX, const CBLAS_INT offsetX ); #ifdef __cplusplus } diff --git a/lib/node_modules/@stdlib/blas/ext/base/dasumpw/lib/dasumpw.js b/lib/node_modules/@stdlib/blas/ext/base/dasumpw/lib/dasumpw.js index 6b65e829e62f..45c8314cfe21 100644 --- a/lib/node_modules/@stdlib/blas/ext/base/dasumpw/lib/dasumpw.js +++ b/lib/node_modules/@stdlib/blas/ext/base/dasumpw/lib/dasumpw.js @@ -20,8 +20,8 @@ // MODULES // -var abs = require( '@stdlib/math/base/special/abs' ); -var sum = require( './ndarray.js' ); +var stride2offset = require( '@stdlib/strided/base/stride2offset' ); +var ndarray = require( './ndarray.js' ); // MAIN // @@ -39,7 +39,7 @@ var sum = require( './ndarray.js' ); * * @param {PositiveInteger} N - number of indexed elements * @param {Float64Array} x - input array -* @param {integer} stride - stride length +* @param {integer} strideX - stride length * @returns {number} sum * * @example @@ -51,32 +51,8 @@ var sum = require( './ndarray.js' ); * var v = dasumpw( N, x, 1 ); * // returns 5.0 */ -function dasumpw( N, x, stride ) { - var ix; - var s; - var i; - - if ( N <= 0 ) { - return 0.0; - } - if ( N === 1 || stride === 0 ) { - return abs( x[ 0 ] ); - } - if ( stride < 0 ) { - ix = (1-N) * stride; - } else { - ix = 0; - } - if ( N < 8 ) { - // Use simple summation... - s = 0.0; - for ( i = 0; i < N; i++ ) { - s += abs( x[ ix ] ); - ix += stride; - } - return s; - } - return sum( N, x, stride, ix ); +function dasumpw( N, x, strideX ) { + return ndarray( N, x, strideX, stride2offset( N, strideX ) ); } diff --git a/lib/node_modules/@stdlib/blas/ext/base/dasumpw/lib/dasumpw.native.js b/lib/node_modules/@stdlib/blas/ext/base/dasumpw/lib/dasumpw.native.js index 315bd7d0f8c5..29556c30cd9b 100644 --- a/lib/node_modules/@stdlib/blas/ext/base/dasumpw/lib/dasumpw.native.js +++ b/lib/node_modules/@stdlib/blas/ext/base/dasumpw/lib/dasumpw.native.js @@ -30,7 +30,7 @@ var addon = require( './../src/addon.node' ); * * @param {PositiveInteger} N - number of indexed elements * @param {Float64Array} x - input array -* @param {integer} stride - stride length +* @param {integer} strideX - stride length * @returns {number} sum * * @example @@ -42,8 +42,8 @@ var addon = require( './../src/addon.node' ); * var v = dasumpw( N, x, 1 ); * // returns 5.0 */ -function dasumpw( N, x, stride ) { - return addon( N, x, stride ); +function dasumpw( N, x, strideX ) { + return addon( N, x, strideX ); } diff --git a/lib/node_modules/@stdlib/blas/ext/base/dasumpw/lib/ndarray.js b/lib/node_modules/@stdlib/blas/ext/base/dasumpw/lib/ndarray.js index 35131a5263d7..e353eb80e6ad 100644 --- a/lib/node_modules/@stdlib/blas/ext/base/dasumpw/lib/ndarray.js +++ b/lib/node_modules/@stdlib/blas/ext/base/dasumpw/lib/ndarray.js @@ -45,8 +45,8 @@ var BLOCKSIZE = 128; * * @param {PositiveInteger} N - number of indexed elements * @param {Float64Array} x - input array -* @param {integer} stride - stride length -* @param {NonNegativeInteger} offset - starting index +* @param {integer} strideX - stride length +* @param {NonNegativeInteger} offsetX - starting index * @returns {number} sum * * @example @@ -57,7 +57,7 @@ var BLOCKSIZE = 128; * var v = dasumpw( 4, x, 2, 1 ); * // returns 9.0 */ -function dasumpw( N, x, stride, offset ) { +function dasumpw( N, x, strideX, offsetX ) { var ix; var s0; var s1; @@ -75,42 +75,42 @@ function dasumpw( N, x, stride, offset ) { if ( N <= 0 ) { return 0.0; } - if ( N === 1 || stride === 0 ) { - return abs( x[ offset ] ); + if ( N === 1 || strideX === 0 ) { + return abs( x[ offsetX ] ); } - ix = offset; + ix = offsetX; if ( N < 8 ) { // Use simple summation... s = 0.0; for ( i = 0; i < N; i++ ) { s += abs( x[ ix ] ); - ix += stride; + ix += strideX; } return s; } if ( N <= BLOCKSIZE ) { // Sum a block with 8 accumulators (by loop unrolling, we lower the effective blocksize to 16)... s0 = abs( x[ ix ] ); - s1 = abs( x[ ix+stride ] ); - s2 = abs( x[ ix+(2*stride) ] ); - s3 = abs( x[ ix+(3*stride) ] ); - s4 = abs( x[ ix+(4*stride) ] ); - s5 = abs( x[ ix+(5*stride) ] ); - s6 = abs( x[ ix+(6*stride) ] ); - s7 = abs( x[ ix+(7*stride) ] ); - ix += 8 * stride; + s1 = abs( x[ ix+strideX ] ); + s2 = abs( x[ ix+(2*strideX) ] ); + s3 = abs( x[ ix+(3*strideX) ] ); + s4 = abs( x[ ix+(4*strideX) ] ); + s5 = abs( x[ ix+(5*strideX) ] ); + s6 = abs( x[ ix+(6*strideX) ] ); + s7 = abs( x[ ix+(7*strideX) ] ); + ix += 8 * strideX; M = N % 8; for ( i = 8; i < N-M; i += 8 ) { s0 += abs( x[ ix ] ); - s1 += abs( x[ ix+stride ] ); - s2 += abs( x[ ix+(2*stride) ] ); - s3 += abs( x[ ix+(3*stride) ] ); - s4 += abs( x[ ix+(4*stride) ] ); - s5 += abs( x[ ix+(5*stride) ] ); - s6 += abs( x[ ix+(6*stride) ] ); - s7 += abs( x[ ix+(7*stride) ] ); - ix += 8 * stride; + s1 += abs( x[ ix+strideX ] ); + s2 += abs( x[ ix+(2*strideX) ] ); + s3 += abs( x[ ix+(3*strideX) ] ); + s4 += abs( x[ ix+(4*strideX) ] ); + s5 += abs( x[ ix+(5*strideX) ] ); + s6 += abs( x[ ix+(6*strideX) ] ); + s7 += abs( x[ ix+(7*strideX) ] ); + ix += 8 * strideX; } // Pairwise sum the accumulators: s = ((s0+s1) + (s2+s3)) + ((s4+s5) + (s6+s7)); @@ -118,14 +118,14 @@ function dasumpw( N, x, stride, offset ) { // Clean-up loop... for ( i; i < N; i++ ) { s += abs( x[ ix ] ); - ix += stride; + ix += strideX; } return s; } // Recurse by dividing by two, but avoiding non-multiples of unroll factor... n = floor( N/2 ); n -= n % 8; - return dasumpw( n, x, stride, ix ) + dasumpw( N-n, x, stride, ix+(n*stride) ); // eslint-disable-line max-len + return dasumpw( n, x, strideX, ix ) + dasumpw( N-n, x, strideX, ix+(n*strideX) ); // eslint-disable-line max-len } diff --git a/lib/node_modules/@stdlib/blas/ext/base/dasumpw/lib/ndarray.native.js b/lib/node_modules/@stdlib/blas/ext/base/dasumpw/lib/ndarray.native.js index c30bcf79ffcf..b2c3a5465987 100644 --- a/lib/node_modules/@stdlib/blas/ext/base/dasumpw/lib/ndarray.native.js +++ b/lib/node_modules/@stdlib/blas/ext/base/dasumpw/lib/ndarray.native.js @@ -20,9 +20,7 @@ // MODULES // -var minViewBufferIndex = require( '@stdlib/strided/base/min-view-buffer-index' ); -var offsetView = require( '@stdlib/strided/base/offset-view' ); -var addon = require( './dasumpw.native.js' ); +var addon = require( './../src/addon.node' ); // MAIN // @@ -32,8 +30,8 @@ var addon = require( './dasumpw.native.js' ); * * @param {PositiveInteger} N - number of indexed elements * @param {Float64Array} x - input array -* @param {integer} stride - stride length -* @param {NonNegativeInteger} offset - starting index +* @param {integer} strideX - stride length +* @param {NonNegativeInteger} offsetX - starting index * @returns {number} sum * * @example @@ -44,11 +42,8 @@ var addon = require( './dasumpw.native.js' ); * var v = dasumpw( 4, x, 2, 1 ); * // returns 9.0 */ -function dasumpw( N, x, stride, offset ) { - var view; - offset = minViewBufferIndex( N, stride, offset ); - view = offsetView( x, offset ); - return addon( N, view, stride ); +function dasumpw( N, x, strideX, offsetX ) { + return addon.ndarray( N, x, strideX, offsetX ); } diff --git a/lib/node_modules/@stdlib/blas/ext/base/dasumpw/manifest.json b/lib/node_modules/@stdlib/blas/ext/base/dasumpw/manifest.json index b2610eaf8e38..dd4b3af0d2d9 100644 --- a/lib/node_modules/@stdlib/blas/ext/base/dasumpw/manifest.json +++ b/lib/node_modules/@stdlib/blas/ext/base/dasumpw/manifest.json @@ -28,49 +28,55 @@ { "task": "build", "src": [ - "./src/dasumpw.c" + "./src/main.c" ], "include": [ "./include" ], - "libraries": [ - "-lm" - ], + "libraries": [], "libpath": [], "dependencies": [ + "@stdlib/math/base/special/abs", + "@stdlib/blas/base/shared", + "@stdlib/strided/base/stride2offset", "@stdlib/napi/export", "@stdlib/napi/argv", "@stdlib/napi/argv-int64", - "@stdlib/napi/argv-strided-float64array" + "@stdlib/napi/argv-strided-float64array", + "@stdlib/napi/create-double" ] }, { "task": "benchmark", "src": [ - "./src/dasumpw.c" + "./src/main.c" ], "include": [ "./include" ], - "libraries": [ - "-lm" - ], + "libraries": [], "libpath": [], - "dependencies": [] + "dependencies": [ + "@stdlib/math/base/special/abs", + "@stdlib/blas/base/shared", + "@stdlib/strided/base/stride2offset" + ] }, { "task": "examples", "src": [ - "./src/dasumpw.c" + "./src/main.c" ], "include": [ "./include" ], - "libraries": [ - "-lm" - ], + "libraries": [], "libpath": [], - "dependencies": [] + "dependencies": [ + "@stdlib/math/base/special/abs", + "@stdlib/blas/base/shared", + "@stdlib/strided/base/stride2offset" + ] } ] } diff --git a/lib/node_modules/@stdlib/blas/ext/base/dasumpw/src/addon.c b/lib/node_modules/@stdlib/blas/ext/base/dasumpw/src/addon.c index 21cdff485dc7..2363a6b25a53 100644 --- a/lib/node_modules/@stdlib/blas/ext/base/dasumpw/src/addon.c +++ b/lib/node_modules/@stdlib/blas/ext/base/dasumpw/src/addon.c @@ -17,12 +17,13 @@ */ #include "stdlib/blas/ext/base/dasumpw.h" +#include "stdlib/blas/base/shared.h" #include "stdlib/napi/export.h" #include "stdlib/napi/argv.h" #include "stdlib/napi/argv_int64.h" #include "stdlib/napi/argv_strided_float64array.h" +#include "stdlib/napi/create_double.h" #include -#include /** * Receives JavaScript callback invocation data. @@ -36,12 +37,25 @@ static napi_value addon( napi_env env, napi_callback_info info ) { STDLIB_NAPI_ARGV_INT64( env, N, argv, 0 ); STDLIB_NAPI_ARGV_INT64( env, strideX, argv, 2 ); STDLIB_NAPI_ARGV_STRIDED_FLOAT64ARRAY( env, X, N, strideX, argv, 1 ); + STDLIB_NAPI_CREATE_DOUBLE( env, API_SUFFIX(stdlib_strided_dasumpw)( N, X, strideX ), v ); + return v; +} - napi_value v; - napi_status status = napi_create_double( env, stdlib_strided_dasumpw( N, X, strideX ), &v ); - assert( status == napi_ok ); - +/** +* Receives JavaScript callback invocation data. +* +* @param env environment under which the function is invoked +* @param info callback data +* @return Node-API value +*/ +static napi_value addon_method( napi_env env, napi_callback_info info ) { + STDLIB_NAPI_ARGV( env, info, argv, argc, 4 ); + STDLIB_NAPI_ARGV_INT64( env, N, argv, 0 ); + STDLIB_NAPI_ARGV_INT64( env, strideX, argv, 2 ); + STDLIB_NAPI_ARGV_INT64( env, offsetX, argv, 3 ); + STDLIB_NAPI_ARGV_STRIDED_FLOAT64ARRAY( env, X, N, strideX, argv, 1 ); + STDLIB_NAPI_CREATE_DOUBLE( env, API_SUFFIX(stdlib_strided_dasumpw_ndarray)( N, X, strideX, offsetX ), v ); return v; } -STDLIB_NAPI_MODULE_EXPORT_FCN( addon ) +STDLIB_NAPI_MODULE_EXPORT_FCN_WITH_METHOD( addon, "ndarray", addon_method ) diff --git a/lib/node_modules/@stdlib/blas/ext/base/dasumpw/src/dasumpw.c b/lib/node_modules/@stdlib/blas/ext/base/dasumpw/src/dasumpw.c deleted file mode 100644 index 80768518a051..000000000000 --- a/lib/node_modules/@stdlib/blas/ext/base/dasumpw/src/dasumpw.c +++ /dev/null @@ -1,122 +0,0 @@ -/** -* @license Apache-2.0 -* -* Copyright (c) 2020 The Stdlib Authors. -* -* Licensed under the Apache License, Version 2.0 (the "License"); -* you may not use this file except in compliance with the License. -* You may obtain a copy of the License at -* -* http://www.apache.org/licenses/LICENSE-2.0 -* -* Unless required by applicable law or agreed to in writing, software -* distributed under the License is distributed on an "AS IS" BASIS, -* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. -* See the License for the specific language governing permissions and -* limitations under the License. -*/ - -#include "stdlib/blas/ext/base/dasumpw.h" -#include -#include - -/** -* Computes the sum of absolute values (L1 norm) of double-precision floating-point strided array elements using pairwise summation. -* -* ## Method -* -* - This implementation uses pairwise summation, which accrues rounding error `O(log2 N)` instead of `O(N)`. The recursion depth is also `O(log2 N)`. -* -* ## References -* -* - Higham, Nicholas J. 1993. "The Accuracy of Floating Point Summation." _SIAM Journal on Scientific Computing_ 14 (4): 783–99. doi:[10.1137/0914050](https://doi.org/10.1137/0914050). -* -* @param N number of indexed elements -* @param X input array -* @param stride stride length -* @return output value -*/ -double stdlib_strided_dasumpw( const int64_t N, const double *X, const int64_t stride ) { - double *xp1; - double *xp2; - double sum; - int64_t ix; - int64_t M; - int64_t n; - int64_t i; - double s0; - double s1; - double s2; - double s3; - double s4; - double s5; - double s6; - double s7; - - if ( N <= 0 ) { - return 0.0; - } - if ( N == 1 || stride == 0 ) { - return fabs( X[ 0 ] ); - } - if ( stride < 0 ) { - ix = (1-N) * stride; - } else { - ix = 0; - } - if ( N < 8 ) { - // Use simple summation... - sum = 0.0; - for ( i = 0; i < N; i++ ) { - sum += fabs( X[ ix ] ); - ix += stride; - } - return sum; - } - // Blocksize for pairwise summation: 128 (NOTE: decreasing the blocksize decreases rounding error as more pairs are summed, but also decreases performance. Because the inner loop is unrolled eight times, the blocksize is effectively `16`.) - if ( N <= 128 ) { - // Sum a block with 8 accumulators (by loop unrolling, we lower the effective blocksize to 16)... - s0 = fabs( X[ ix ] ); - s1 = fabs( X[ ix+stride ] ); - s2 = fabs( X[ ix+(2*stride) ] ); - s3 = fabs( X[ ix+(3*stride) ] ); - s4 = fabs( X[ ix+(4*stride) ] ); - s5 = fabs( X[ ix+(5*stride) ] ); - s6 = fabs( X[ ix+(6*stride) ] ); - s7 = fabs( X[ ix+(7*stride) ] ); - ix += 8 * stride; - - M = N % 8; - for ( i = 8; i < N-M; i += 8 ) { - s0 += fabs( X[ ix ] ); - s1 += fabs( X[ ix+stride ] ); - s2 += fabs( X[ ix+(2*stride) ] ); - s3 += fabs( X[ ix+(3*stride) ] ); - s4 += fabs( X[ ix+(4*stride) ] ); - s5 += fabs( X[ ix+(5*stride) ] ); - s6 += fabs( X[ ix+(6*stride) ] ); - s7 += fabs( X[ ix+(7*stride) ] ); - ix += 8 * stride; - } - // Pairwise sum the accumulators: - sum = ((s0+s1) + (s2+s3)) + ((s4+s5) + (s6+s7)); - - // Clean-up loop... - for (; i < N; i++ ) { - sum += fabs( X[ ix ] ); - ix += stride; - } - return sum; - } - // Recurse by dividing by two, but avoiding non-multiples of unroll factor... - n = N / 2; - n -= n % 8; - if ( stride < 0 ) { - xp1 = (double *)X + ( (n-N)*stride ); - xp2 = (double *)X; - } else { - xp1 = (double *)X; - xp2 = (double *)X + ( n*stride ); - } - return stdlib_strided_dasumpw( n, xp1, stride ) + stdlib_strided_dasumpw( N-n, xp2, stride ); -} diff --git a/lib/node_modules/@stdlib/blas/ext/base/dasumpw/src/main.c b/lib/node_modules/@stdlib/blas/ext/base/dasumpw/src/main.c new file mode 100644 index 000000000000..50ff029d1722 --- /dev/null +++ b/lib/node_modules/@stdlib/blas/ext/base/dasumpw/src/main.c @@ -0,0 +1,132 @@ +/** +* @license Apache-2.0 +* +* Copyright (c) 2020 The Stdlib Authors. +* +* Licensed under the Apache License, Version 2.0 (the "License"); +* you may not use this file except in compliance with the License. +* You may obtain a copy of the License at +* +* http://www.apache.org/licenses/LICENSE-2.0 +* +* Unless required by applicable law or agreed to in writing, software +* distributed under the License is distributed on an "AS IS" BASIS, +* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +* See the License for the specific language governing permissions and +* limitations under the License. +*/ + +#include "stdlib/blas/ext/base/dasumpw.h" +#include "stdlib/blas/base/shared.h" +#include "stdlib/strided/base/stride2offset.h" +#include "stdlib/math/base/special/abs.h" + +/** +* Computes the sum of absolute values (L1 norm) of double-precision floating-point strided array elements using pairwise summation. +* +* ## Method +* +* - This implementation uses pairwise summation, which accrues rounding error `O(log2 N)` instead of `O(N)`. The recursion depth is also `O(log2 N)`. +* +* ## References +* +* - Higham, Nicholas J. 1993. "The Accuracy of Floating Point Summation." _SIAM Journal on Scientific Computing_ 14 (4): 783–99. doi:[10.1137/0914050](https://doi.org/10.1137/0914050). +* +* @param N number of indexed elements +* @param X input array +* @param strideX stride length +* @return output value +*/ +double API_SUFFIX(stdlib_strided_dasumpw)( const CBLAS_INT N, const double *X, const CBLAS_INT strideX ) { + CBLAS_INT ox = stdlib_strided_stride2offset( N, strideX ); + return API_SUFFIX(stdlib_strided_dasumpw_ndarray)( N, X, strideX, ox ); +} + +/** +* Computes the sum of absolute values (L1 norm) of double-precision floating-point strided array elements using pairwise summation and alternative indexing semantics. +* +* ## Method +* +* - This implementation uses pairwise summation, which accrues rounding error `O(log2 N)` instead of `O(N)`. The recursion depth is also `O(log2 N)`. +* +* ## References +* +* - Higham, Nicholas J. 1993. "The Accuracy of Floating Point Summation." _SIAM Journal on Scientific Computing_ 14 (4): 783–99. doi:[10.1137/0914050](https://doi.org/10.1137/0914050). +* +* @param N number of indexed elements +* @param X input array +* @param strideX index increment +* @param offsetX starting index +* @return output value +*/ +double API_SUFFIX(stdlib_strided_dasumpw_ndarray)( const CBLAS_INT N, const double *X, const CBLAS_INT strideX, const CBLAS_INT offsetX ) { + double sum; + CBLAS_INT ix; + CBLAS_INT M; + CBLAS_INT n; + CBLAS_INT i; + double s0; + double s1; + double s2; + double s3; + double s4; + double s5; + double s6; + double s7; + + if ( N <= 0 ) { + return 0.0; + } + if ( N == 1 || strideX == 0 ) { + return stdlib_base_abs( X[ 0 ] ); + } + ix = offsetX; + if ( N < 8 ) { + // Use simple summation... + sum = 0.0; + for ( i = 0; i < N; i++ ) { + sum += stdlib_base_abs( X[ ix ] ); + ix += strideX; + } + return sum; + } + // Blocksize for pairwise summation: 128 (NOTE: decreasing the blocksize decreases rounding error as more pairs are summed, but also decreases performance. Because the inner loop is unrolled eight times, the blocksize is effectively `16`.) + if ( N <= 128 ) { + // Sum a block with 8 accumulators (by loop unrolling, we lower the effective blocksize to 16)... + s0 = stdlib_base_abs( X[ ix ] ); + s1 = stdlib_base_abs( X[ ix+strideX ] ); + s2 = stdlib_base_abs( X[ ix+(2*strideX) ] ); + s3 = stdlib_base_abs( X[ ix+(3*strideX) ] ); + s4 = stdlib_base_abs( X[ ix+(4*strideX) ] ); + s5 = stdlib_base_abs( X[ ix+(5*strideX) ] ); + s6 = stdlib_base_abs( X[ ix+(6*strideX) ] ); + s7 = stdlib_base_abs( X[ ix+(7*strideX) ] ); + ix += 8 * strideX; + + M = N % 8; + for ( i = 8; i < N-M; i += 8 ) { + s0 += stdlib_base_abs( X[ ix ] ); + s1 += stdlib_base_abs( X[ ix+strideX ] ); + s2 += stdlib_base_abs( X[ ix+(2*strideX) ] ); + s3 += stdlib_base_abs( X[ ix+(3*strideX) ] ); + s4 += stdlib_base_abs( X[ ix+(4*strideX) ] ); + s5 += stdlib_base_abs( X[ ix+(5*strideX) ] ); + s6 += stdlib_base_abs( X[ ix+(6*strideX) ] ); + s7 += stdlib_base_abs( X[ ix+(7*strideX) ] ); + ix += 8 * strideX; + } + // Pairwise sum the accumulators: + sum = ((s0+s1) + (s2+s3)) + ((s4+s5) + (s6+s7)); + + // Clean-up loop... + for (; i < N; i++ ) { + sum += stdlib_base_abs( X[ ix ] ); + ix += strideX; + } + return sum; + } + // Recurse by dividing by two, but avoiding non-multiples of unroll factor... + n = N / 2; + n -= n % 8; + return API_SUFFIX(stdlib_strided_dasumpw_ndarray)( n, X, strideX, ix ) + API_SUFFIX(stdlib_strided_dasumpw_ndarray)( N-n, X, strideX, ix+(n*strideX) ); +}