diff --git a/bin/NLMSvariants.c b/bin/NLMSvariants.c new file mode 100644 index 0000000..af1b6d9 --- /dev/null +++ b/bin/NLMSvariants.c @@ -0,0 +1,287 @@ +// +// +// NLMSvariants.c +// +// Created by FBRDNLMS on 26.04.18. +// Copyright © 2018 FBRDNLMS. All rights reserved. +// + +#include +#include +#include +#include +#include + +#define M 1000 +#define tracking 40 //Count of weights +#define learnrate 1.0 +#define WGHTS 1 +#define WGHTSFTR 2 +#define RES 3 +#define DRCTPRD 4 + +double x[] ={0}; +double _x[M] = {0}; +double w [M][M]={{0},{0}}; + +const char *fileName(int id); +double r2(void); +double rnd(void); +double sum_array(double x[], int length); +void direkterVorgaenger(void); +void lokalerMittelWert(void); + + +int main(int argc, char **argv ) { + + //init test_array, fill in weights by random + int i; + // srand(NULL); + for (i = 0; i < M; i++) { + _x[i] += ((255.0 / M) * i); + for (int k = 1; k < M; k++) + { + w[k][i] = rnd(); + } + } + + // save plain test_array before math magic happened + FILE *fp0 = fopen(fileName(WGHTS),"wb+"); + for (i = 0; i < tracking; i++){ + for ( int k = 1; k < tracking; k++ ){ + fprintf(fp0, "[%f][%f] %.2f\n", k, i, w[k][i]); + } + } + fclose(fp0); + + + // math magic + direkterVorgaenger(); + + + // save test_array after math magic happened + FILE *fp1 = fopen(fileName(WGHTSFTR),"wb+"); + for (i = 0; i < tracking; i++) { + for (int k = 1; k < tracking; k++) { + fprintf(fp1, "[%f][%f] %.2f\n", k,i, w[k][i]); + } + + } + fclose(fp1); + + getchar(); + +} + + +/* + + =================================== + + lokalerMittelwert() + + + Variant (1/3), + substract local mean + + =================================== +*/ + +void lokalerMittelWert() { + + double xError[M]; // includes e(n) + memset(xError, 0, M);// initialize xError-array with Zero + int xCount = 0; // runtime var + int i; + + + for (xCount = 1; xCount < M; xCount++){ // x_cout can not be zero + + //double xPartArray[xCount]; //includes all values at the size of runtime var + + double xMean = (xCount > 0) ? ( sum_array(_x, xCount) / xCount) : 0; + + double xPredicted = 0.0; + double xActual = _x[xCount + 1]; + + for ( i = 1; i < xCount; i++ ){ //get predicted value + xPredicted += (w[i][xCount] * (_x[xCount - i] - xMean)) ; + } + + xPredicted += xMean; + xError [xCount] = xActual - xPredicted; + + double xSquared = 0.0; + + for ( i = 1; i < xCount; i++ ){ //get x squared + xSquared =+ pow(_x[xCount-i],2); + } + + for ( i - 1; i < xCount; i++ ){ //update weights + w[i][xCount+1] = w[i][xCount] + learnrate * xError[xCount] * (_x[xCount - i] / xSquared); + } + } + + int xErrorLength = sizeof(xError) / sizeof(xError[0]); + double mean = sum_array(xError, xErrorLength) / M; + double deviation = 0.0; + + // Mean square + for( i = 0; i < M-1; i++ ){ + deviation += pow( xError[i], 2 ); + } + deviation /= xErrorLength; + + + // write in file + FILE *fp2 = fopen(fileName(RES), "wb+"); + fprintf(fp2, "quadr. Varianz(x_error): {%f}\nMittelwert:(x_error): {%f}\n\n", deviation, mean); + fclose(fp2); + +} + + +/* + + =================================== + + direkterVorgaenger() + + + Variant (2/3), + substract direct predecessor + + =================================== +*/ + +void direkterVorgaenger() { + + double xError [M]; + int xCount = 0, i; + + // File handling + FILE *fp3 = fopen(fileName(DRCTPRD), "wb+"); + + for ( xCount = 1; xCount < M; xCount++ ){ + double xPredicted = 0.0; + double xActual = _x[xCount+1]; + + for ( i = 1; i < xCount; i++ ){ + xPredicted += ( w[i][xCount] * ( _x[xCount - i] - _x[xCount - i - 1])); + } + + xPredicted += _x[xCount-1]; + xError[xCount] = xActual - xPredicted; + + fprintf(fp3, "{%d}.\txPredicted{%f}\txActual{%f}\txError{%f}\n", xCount, xPredicted, xActual, xError[xCount]); + + + //get x squared + double xSquared = 0; + for ( i = 1; i < xCount; i++ ){ + xSquared += pow( _x[xCount - i] - _x[xCount - i - 1], 2); // substract direct predecessor + } + + for ( i = 1; i < xCount; i++){ + w[i][xCount+1] = w[i][xCount] + learnrate * xError[xCount] * ( ( _x[xCount - i - 1] ) / xSquared ); + } + } + int xErrorLength = sizeof(xError) / sizeof(xError[0]); + double mean = sum_array(xError, xErrorLength) / xErrorLength; + double deviation = 0.0; + + for ( i = 0; i < xErrorLength -1; i++ ){ + deviation += pow( xError[i] - mean, 2); + } + + + fprintf(fp3, "{%d}.\tLeast Mean Squared{%f}\tMean{%f}\n\n", xCount, deviation, mean); + fclose(fp3); +} + + +/* + + =================================== + + fileName() + + + generates filename with date for + logging purposes + + =================================== +*/ + +const char *fileName(int id){ + + static char* suffix[] = {"_weights.txt","_weights_after.txt", "_results.txt", "direct_predecessor.txt"}; + char *date; + time_t now; + now = time(NULL); + strftime(date, 20, "%Y-%m-%d_%H_%M_%S", localtime(&now)); + strcat(suffix[id], date); + return suffix[id]; +} + + +/* + + =================================== + + sum_array + + + sum of all elements in x + within a defined length + + =================================== + + */ + +double sum_array(double x[], int length){ + //int length = 0; + int i = 0; + double sum = 0.0; + //length = sizeof(x)/sizeof(x[0]); + for (i=0; i< length; i++){ + sum += x[i]; + } + return sum; +} + + +/* + + =================================== + + r2() + + returns a double value between + 0 and 1 + + =================================== + +*/ +double r2() +{ + return((rand() % 10000) / 10000.0); +} + +/* + + =================================== + + int rnd() + + fills a double variable with + random value and returns it + + =================================== + +*/ +double rnd() +{ + double rndmval= r2(); + return rndmval; +} diff --git a/bin/main.c b/bin/main.c deleted file mode 100644 index abb57ba..0000000 --- a/bin/main.c +++ /dev/null @@ -1,3 +0,0 @@ -#include -#include -#include \ No newline at end of file