321a035f16d6 — Chris Cannam 2 years ago
More meaningful timing code
1 files changed, 155 insertions(+), 177 deletions(-)

M test/Timings.cpp
M test/Timings.cpp +155 -177
@@ 1,6 1,7 @@ 
 /* -*- c-basic-offset: 4 indent-tabs-mode: nil -*-  vi:set ts=8 sts=4 sw=4: */
 
 #include "bqvec/VectorOpsComplex.h"
+#include "bqvec/Allocators.h"
 
 #include <iostream>
 #include <cstdlib>

          
@@ 10,8 11,6 @@ 
 using namespace std;
 using namespace breakfastquay;
 
-//!!! This is nonsense. TODO: Replace it with sense.
-
 #ifdef _WIN32
 #define drand48() (-1+2*((float)rand())/RAND_MAX)
 #endif

          
@@ 21,11 20,12 @@ testMultiply()
 {
     cerr << "testVectorOps: testing v_multiply complex" << endl;
 
+    const int iterations = 50000;
     const int N = 1024;
-    //!!! todo: use aligned allocate(), otherwise results will vary randomly
-    bq_complex_t target[N];
-    bq_complex_t src1[N];
-    bq_complex_t src2[N];
+
+    bq_complex_t *target = allocate<bq_complex_t>(N);
+    bq_complex_t *src1 = allocate<bq_complex_t>(N);
+    bq_complex_t *src2 = allocate<bq_complex_t>(N);
 
     for (int i = 0; i < N; ++i) {
 	src1[i].re = drand48();

          
@@ 34,60 34,55 @@ testMultiply()
 	src2[i].im = drand48();
     }
 
-    double mean, first, last, total = 0;
-    for (int i = 0; i < N; ++i) {
-        bq_complex_t result;
-        c_multiply(result, src1[i], src2[i]);
-	if (i == 0) first = result.re;
-	if (i == N-1) last = result.im;
-	total += result.re;
-	total += result.im;
-    }
-    mean = total / (N*2);
-    cerr << "Naive method: mean = " << mean << ", first = " << first
-	 << ", last = " << last << endl;
-
-    v_multiply_to(target, src1, src2, N);
-    total = 0;
-
-    for (int i = 0; i < N; ++i) {
-	if (i == 0) first = target[i].re;
-	if (i == N-1) last = target[i].im;
-	total += target[i].re;
-	total += target[i].im;
-    }
-    mean = total / (N*2);
-    cerr << "v_multiply: mean = " << mean << ", first = " << first
-	 << ", last = " << last << endl;
-
-    int iterations = 50000;
-//    cerr << "Iterations: " << iterations << endl;
-	
-//    cerr << "CLOCKS_PER_SEC = " << CLOCKS_PER_SEC << endl;
     float divisor = float(CLOCKS_PER_SEC) / 1000.f;
-
     clock_t start = clock();
 
+    double first, last, total = 0;
     for (int j = 0; j < iterations; ++j) {
-	for (int i = 0; i < N; ++i) {
-            c_multiply(target[i], src1[i], src2[i]);
-	}
+        for (int i = 0; i < N; ++i) {
+            bq_complex_t result;
+            c_multiply(result, src1[i], src2[i]);
+            if (i == 0) first = result.re;
+            if (i == N-1) last = result.im;
+            total += result.re;
+            total += result.im;
+            total += j;
+        }
     }
-    
+
     clock_t end = clock();
 
-    cerr << "Time for naive method: " << float(end - start)/divisor << endl;
+    cerr << "repeated c_multiply: first = " << first << ", last = " << last
+         << ", total = " << total << endl;
+    cerr << "time for repeated c_multiply: " << float(end - start)/divisor
+         << endl;
 
     start = clock();
 
+    first = last = total = 0;
+
     for (int j = 0; j < iterations; ++j) {
         v_multiply_to(target, src1, src2, N);
+        for (int i = 0; i < N; ++i) {
+            if (i == 0) first = target[i].re;
+            if (i == N-1) last = target[i].im;
+            total += target[i].re;
+            total += target[i].im;
+            total += j;
+        }
     }
-    
+
     end = clock();
 
-    cerr << "Time for v_multiply: " << float(end - start)/divisor << endl;
+    cerr << "v_multiply_to: first = " << first << ", last = " << last
+         << ", total = " << total << endl;
+    cerr << "time for v_multiply_to: " << float(end - start)/divisor << endl;
 
+    deallocate(target);
+    deallocate(src1);
+    deallocate(src2);
+    
+    cerr << endl;
     return true;
 }
 

          
@@ 96,72 91,68 @@ testPolarToCart()
 {
     cerr << "testVectorOps: testing v_polar_to_cartesian" << endl;
 
+    const int iterations = 50000;
     const int N = 1024;
-    bq_complex_t target[N];
-    bq_complex_element_t mag[N];
-    bq_complex_element_t phase[N];
+    
+    bq_complex_t *target = allocate<bq_complex_t>(N);
+    bq_complex_element_t *mag = allocate<bq_complex_element_t>(N);
+    bq_complex_element_t *phase = allocate<bq_complex_element_t>(N);
 
     for (int i = 0; i < N; ++i) {
 	mag[i] = drand48();
 	phase[i] = (drand48() * M_PI * 2) - M_PI;
     }
 
-    double mean, first, last, total = 0;
-    for (int i = 0; i < N; ++i) {
-	double real = mag[i] * cos(phase[i]);
-	double imag = mag[i] * sin(phase[i]);
-	if (i == 0) first = real;
-	if (i == N-1) last = imag;
-	total += real;
-	total += imag;
-    }
-    mean = total / (N*2);
-    cerr << "Naive method: mean = " << mean << ", first = " << first
-	 << ", last = " << last << endl;
-
-    v_polar_to_cartesian(target, mag, phase, N);
-
-    total = 0;
-
-    for (int i = 0; i < N; ++i) {
-	if (i == 0) first = target[i].re;
-	if (i == N-1) last = target[i].im;
-	total += target[i].re;
-	total += target[i].im;
-    }
-    mean = total / (N*2);
-    cerr << "v_polar_to_cartesian: mean = " << mean << ", first = " << first
-	 << ", last = " << last << endl;
-
-    int iterations = 10000;
-//    cerr << "Iterations: " << iterations << endl;
-	
-//    cerr << "CLOCKS_PER_SEC = " << CLOCKS_PER_SEC << endl;
     float divisor = float(CLOCKS_PER_SEC) / 1000.f;
-
     clock_t start = clock();
 
+    double first, last, total = 0;
     for (int j = 0; j < iterations; ++j) {
-	for (int i = 0; i < N; ++i) {
-	    target[i].re = mag[i] * cos(phase[i]);
-	    target[i].im = mag[i] * sin(phase[i]);
-	}
+        for (int i = 0; i < N; ++i) {
+            double real = mag[i] * cos(phase[i]);
+            double imag = mag[i] * sin(phase[i]);
+            if (i == 0) first = real;
+            if (i == N-1) last = imag;
+            total += real;
+            total += imag;
+            total += j;
+        }
     }
-    
+
     clock_t end = clock();
 
-    cerr << "Time for naive method: " << float(end - start)/divisor << endl;
-
+    cerr << "naive method: first = " << first << ", last = " << last
+         << ", total = " << total << endl;
+    cerr << "time for naive method: " << float(end - start)/divisor
+         << endl;
+    
     start = clock();
 
+    first = last = total = 0;
+    
     for (int j = 0; j < iterations; ++j) {
-	v_polar_to_cartesian(target, mag, phase, N);
+        v_polar_to_cartesian(target, mag, phase, N);
+        for (int i = 0; i < N; ++i) {
+            if (i == 0) first = target[i].re;
+            if (i == N-1) last = target[i].im;
+            total += target[i].re;
+            total += target[i].im;
+            total += j;
+        }
     }
-    
+
     end = clock();
 
-    cerr << "Time for v_polar_to_cartesian: " << float(end - start)/divisor << endl;
-
+    cerr << "v_polar_to_cartesian: first = " << first << ", last = " << last
+         << ", total = " << total << endl;
+    cerr << "time for v_polar_to_cartesian: " << float(end - start)/divisor
+         << endl;
+    
+    deallocate(target);
+    deallocate(mag);
+    deallocate(phase);
+    
+    cerr << endl;
     return true;
 }
 

          
@@ 170,71 161,66 @@ testPolarToCartInterleaved()
 {
     cerr << "testVectorOps: testing v_polar_interleaved_to_cartesian" << endl;
 
+    const int iterations = 50000;
     const int N = 1024;
-    bq_complex_t target[N];
-    bq_complex_element_t source[N*2];
+
+    bq_complex_t *target = allocate<bq_complex_t>(N);
+    bq_complex_element_t *source = allocate<bq_complex_element_t>(N*2);
 
     for (int i = 0; i < N; ++i) {
 	source[i*2] = drand48();
 	source[i*2+1] = (drand48() * M_PI * 2) - M_PI;
     }
 
-    double mean, first, last, total = 0;
-    for (int i = 0; i < N; ++i) {
-	double real = source[i*2] * cos(source[i*2+1]);
-	double imag = source[i*2] * sin(source[i*2+1]);
-	if (i == 0) first = real;
-	if (i == N-1) last = imag;
-	total += real;
-	total += imag;
-    }
-    mean = total / (N*2);
-    cerr << "Naive method: mean = " << mean << ", first = " << first
-	 << ", last = " << last << endl;
-
-    v_polar_interleaved_to_cartesian(target, source, N);
-
-    total = 0;
-
-    for (int i = 0; i < N; ++i) {
-	if (i == 0) first = target[i].re;
-	if (i == N-1) last = target[i].im;
-	total += target[i].re;
-	total += target[i].im;
-    }
-    mean = total / (N*2);
-    cerr << "v_polar_interleaved_to_cartesian: mean = " << mean << ", first = " << first
-	 << ", last = " << last << endl;
-
-    int iterations = 10000;
-//    cerr << "Iterations: " << iterations << endl;
-	
-//    cerr << "CLOCKS_PER_SEC = " << CLOCKS_PER_SEC << endl;
     float divisor = float(CLOCKS_PER_SEC) / 1000.f;
-
     clock_t start = clock();
 
+    double first, last, total = 0;
     for (int j = 0; j < iterations; ++j) {
-	for (int i = 0; i < N; ++i) {
-	    target[i].re = source[i*2] * cos(source[i*2+1]);
-	    target[i].im = source[i*2] * sin(source[i*2+1]);
-	}
+        for (int i = 0; i < N; ++i) {
+            double real = source[i*2] * cos(source[i*2+1]);
+            double imag = source[i*2] * sin(source[i*2+1]);
+            if (i == 0) first = real;
+            if (i == N-1) last = imag;
+            total += real;
+            total += imag;
+            total += j;
+        }
     }
-    
+
     clock_t end = clock();
 
-    cerr << "Time for naive method: " << float(end - start)/divisor << endl;
-
+    cerr << "naive method: first = " << first << ", last = " << last
+         << ", total = " << total << endl;
+    cerr << "time for naive method: " << float(end - start)/divisor
+         << endl;
+    
     start = clock();
 
+    first = last = total = 0;
+    
     for (int j = 0; j < iterations; ++j) {
-	v_polar_interleaved_to_cartesian(target, source, N);
+        v_polar_interleaved_to_cartesian(target, source, N);
+        for (int i = 0; i < N; ++i) {
+            if (i == 0) first = target[i].re;
+            if (i == N-1) last = target[i].im;
+            total += target[i].re;
+            total += target[i].im;
+            total += j;
+        }
     }
-    
+
     end = clock();
 
-    cerr << "Time for v_polar_interleaved_to_cartesian: " << float(end - start)/divisor << endl;
-
+    cerr << "v_polar_interleaved_to_cartesian: first = " << first << ", last = " << last
+         << ", total = " << total << endl;
+    cerr << "time for v_polar_interleaved_to_cartesian: " << float(end - start)/divisor
+         << endl;
+    
+    deallocate(target);
+    deallocate(source);
+    
+    cerr << endl;
     return true;
 }
 

          
@@ 243,77 229,69 @@ testCartToPolar()
 {
     cerr << "testVectorOps: testing v_cartesian_to_polar" << endl;
 
+    const int iterations = 50000;
     const int N = 1024;
-    bq_complex_t source[N];
-    bq_complex_element_t mag[N];
-    bq_complex_element_t phase[N];
+
+    bq_complex_t *source = allocate<bq_complex_t>(N);
+    bq_complex_element_t *mag = allocate<bq_complex_element_t>(N);
+    bq_complex_element_t *phase = allocate<bq_complex_element_t>(N);
 
     for (int i = 0; i < N; ++i) {
         source[i].re = (drand48() * 2.0) - 1.0;
         source[i].im = (drand48() * 2.0) - 1.0;
     }
 
-    double mean, first, last, total = 0;
-    for (int i = 0; i < N; ++i) {
-        double mag = sqrt(source[i].re * source[i].re + source[i].im * source[i].im);
-        double phase = atan2(source[i].im, source[i].re);
-	if (i == 0) first = mag;
-	if (i == N-1) last = phase;
-	total += mag;
-	total += phase;
-    }
-    mean = total / (N*2);
-    cerr << "Naive method: mean = " << mean << ", first = " << first
-	 << ", last = " << last << endl;
-
-    v_cartesian_to_polar(mag, phase, source, N);
-
-    total = 0;
-
-    for (int i = 0; i < N; ++i) {
-	if (i == 0) first = mag[i];
-	if (i == N-1) last = phase[i];
-	total += mag[i];
-	total += phase[i];
-    }
-    mean = total / (N*2);
-    cerr << "v_cartesian_to_polar: mean = " << mean << ", first = " << first
-	 << ", last = " << last << endl;
-
-    int iterations = 10000;
-//    cerr << "Iterations: " << iterations << endl;
-	
-//    cerr << "CLOCKS_PER_SEC = " << CLOCKS_PER_SEC << endl;
     float divisor = float(CLOCKS_PER_SEC) / 1000.f;
-
     clock_t start = clock();
 
+    double first, last, total = 0;
     for (int j = 0; j < iterations; ++j) {
-	for (int i = 0; i < N; ++i) {
-            mag[i] = sqrt(source[i].re * source[i].re + source[i].im * source[i].im);
-            phase[i] = atan2(source[i].im, source[i].re);
-	}
+        for (int i = 0; i < N; ++i) {
+            double mag = sqrt(source[i].re * source[i].re + source[i].im * source[i].im);
+            double phase = atan2(source[i].im, source[i].re);
+            if (i == 0) first = mag;
+            if (i == N-1) last = phase;
+            total += mag;
+            total += j * phase;
+        }
     }
     
     clock_t end = clock();
 
-    cerr << "Time for naive method: " << float(end - start)/divisor << endl;
+    cerr << "naive method: first = " << first << ", last = " << last
+         << ", total = " << total << endl;
+    cerr << "time for naive method: " << float(end - start)/divisor << endl;
 
     start = clock();
 
+    first = last = total = 0;
     for (int j = 0; j < iterations; ++j) {
         v_cartesian_to_polar(mag, phase, source, N);
+        for (int i = 0; i < N; ++i) {
+            if (i == 0) first = mag[i];
+            if (i == N-1) last = phase[i];
+            total += mag[i];
+            total += j * phase[i];
+        }
     }
-    
+
     end = clock();
 
-    cerr << "Time for v_cartesian_to_polar: " << float(end - start)/divisor << endl;
+    cerr << "v_cartesian_to_polar: first = " << first << ", last = " << last
+         << ", total = " << total << endl;
+    cerr << "time for v_cartesian_to_polar: " << float(end - start)/divisor << endl;
 
+    deallocate(source);
+    deallocate(mag);
+    deallocate(phase);
+    
+    cerr << endl;
     return true;
 }
 
 int main(int, char **)
 {
+    cerr << endl;
     if (!testMultiply()) return 1;
     if (!testPolarToCart()) return 1;
     if (!testPolarToCartInterleaved()) return 1;