debugged

Reviewed-on: freudenreichan/info2Praktikum-NeuronalesNetz#3

.vscode/launch.json

@@ -0,0 +1,33 @@
+{
+    // Use IntelliSense to learn about possible attributes.
+    // Hover to view descriptions of existing attributes.
+    // For more information, visit: https://go.microsoft.com/fwlink/?linkid=830387
+    "version": "0.2.0",
+    "configurations": [
+    {
+        "name": "(gdb) Launch",
+        "type": "cppdbg",
+        "request": "launch",
+        "program": "${workspaceFolder}/mnist.exe",
+        "args": ["mnist_test.info2","mnist_model.info2"],
+        "stopAtEntry": false,
+        "cwd": "${fileDirname}",
+        "environment": [],
+        "externalConsole": false,
+        "MIMode": "gdb",
+        "miDebuggerPath": "C:\\msys64\\ucrt64\\bin\\gdb.exe",
+        "setupCommands": [
+            {
+                "description": "Enable pretty-printing for gdb",
+                "text": "-enable-pretty-printing",
+                "ignoreFailures": true
+            },
+            {
+                "description": "Set Disassembly Flavor to Intel",
+                "text": "-gdb-set disassembly-flavor intel",
+                "ignoreFailures": true
+            }
+        ]
+    }
+    ]
+}

imageInput.c

@@ -1,22 +1,97 @@
-#include <stdio.h>
-#include <stdlib.h>
-#include <string.h>
-#include "imageInput.h"
-
-#define BUFFER_SIZE 100
-#define FILE_HEADER_STRING "__info2_image_file_format__"
-
-// TODO Implementieren Sie geeignete Hilfsfunktionen für das Lesen der Bildserie aus einer Datei
-
-// TODO Vervollständigen Sie die Funktion readImages unter Benutzung Ihrer Hilfsfunktionen
-GrayScaleImageSeries *readImages(const char *path)
-{
-    GrayScaleImageSeries *series = NULL;
-    
-    return series;
-}
-
-// TODO Vervollständigen Sie die Funktion clearSeries, welche eine Bildserie vollständig aus dem Speicher freigibt
-void clearSeries(GrayScaleImageSeries *series)
-{
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include "imageInput.h"
+
+#define BUFFER_SIZE 100
+#define FILE_HEADER_STRING "__info2_image_file_format__"
+
+// TODO Implementieren Sie geeignete Hilfsfunktionen für das Lesen der Bildserie aus einer Datei
+
+// TODO Vervollständigen Sie die Funktion readImages unter Benutzung Ihrer Hilfsfunktionen
+GrayScaleImageSeries *readImages(const char *path)
+{
+    FILE *file = fopen(path, "rb");
+    if (file == NULL)
+    {
+        return NULL;
+    }
+
+    // Überprüfe den Header
+    char fileTag[strlen(FILE_HEADER_STRING)];
+    fread(fileTag, sizeof(fileTag[0]), strlen(FILE_HEADER_STRING), file);
+
+    if (strcmp(fileTag, FILE_HEADER_STRING) != 0)
+    {
+        fclose(file);
+        return NULL;
+    }
+
+    // Lese die Metadaten: Anzahl der Bilder, Breite und Höhe
+    unsigned short numberOfImages, width, height;
+    fread(&numberOfImages, sizeof(numberOfImages), 1, file);
+    fread(&width, sizeof(width), 1, file);
+    fread(&height, sizeof(height), 1, file);
+
+    GrayScaleImageSeries *series = (GrayScaleImageSeries *)malloc(sizeof(GrayScaleImageSeries));
+    if (series == NULL)
+    {
+        fclose(file);
+        return NULL;
+    }
+
+    series->count = numberOfImages;
+    series->images = (GrayScaleImage *)malloc(numberOfImages * sizeof(GrayScaleImage));
+    series->labels = (unsigned char *)malloc(numberOfImages * sizeof(unsigned char));
+
+    if (series->images == NULL || series->labels == NULL)
+    {
+        free(series);
+        fclose(file);
+        return NULL;
+    }
+
+    for (int i = 0; i < numberOfImages; i++)
+    {
+        series->images[i].width = width;
+        series->images[i].height = height;
+        series->images[i].buffer = (GrayScalePixelType *)malloc(width * height * sizeof(GrayScalePixelType));
+
+        if (series->images[i].buffer == NULL)
+        {
+            // Fehlerbehandlung: Speicher freigeben, wenn malloc fehlschlägt
+            for (int j = 0; j < i; j++)
+            {
+                free(series->images[j].buffer);
+            }
+            free(series->images);
+            free(series->labels);
+            free(series);
+            fclose(file);
+            return NULL;
+        }
+
+        // Lese die Pixel-Daten und das Label
+        fread(series->images[i].buffer, sizeof(GrayScalePixelType), width * height, file);
+        fread(&series->labels[i], sizeof(unsigned char), 1, file);
+    }
+
+    fclose(file);
+    return series;
+}
+
+// TODO Vervollständigen Sie die Funktion clearSeries, welche eine Bildserie vollständig aus dem Speicher freigibt
+void clearSeries(GrayScaleImageSeries *series)
+{
+    if (series == NULL)
+        return;
+
+    for (int i = 0; i < series->count; i++)
+    {
+        free(series->images[i].buffer); // Speicher für das Bild freigeben
+    }
+
+    free(series->images); // Speicher für die Bild-Array freigeben
+    free(series->labels); // Speicher für die Labels freigeben
+    free(series); // Speicher freigeben
 }

imageInputTests.c

@@ -54,7 +54,7 @@ void test_readImagesReturnsCorrectImageWidth(void)
     GrayScaleImageSeries *series = NULL;
     const unsigned short expectedWidth = 10;
     const char *path = "testFile.info2";
-    prepareImageFile(path, 8, expectedWidth, 2, 1);
+    prepareImageFile(path, expectedWidth, 8, 2, 1);
     series = readImages(path);
     TEST_ASSERT_NOT_NULL(series);
     TEST_ASSERT_NOT_NULL(series->images);
@@ -70,7 +70,7 @@ void test_readImagesReturnsCorrectImageHeight(void)
     GrayScaleImageSeries *series = NULL;
     const unsigned short expectedHeight = 10;
     const char *path = "testFile.info2";
-    prepareImageFile(path, expectedHeight, 8, 2, 1);
+    prepareImageFile(path, 8, expectedHeight, 2, 1);
     series = readImages(path);
     TEST_ASSERT_NOT_NULL(series);
     TEST_ASSERT_NOT_NULL(series->images);

matrix.c

@@ -1,123 +1,154 @@
-#include <stdlib.h>
-#include <string.h>
-#include "matrix.h"
-
-// TODO Matrix-Funktionen implementieren
-
-Matrix createMatrix(unsigned int rows, unsigned int cols)
-{
-    Matrix m = {0, 0, NULL};
-
-    if (rows > 0 && cols > 0)
-    {
-        m.rows = rows;
-        m.cols = cols;
-        m.buffer = malloc(rows * cols * sizeof(int));
-    }
-
-    return m;
-}
-
-void clearMatrix(Matrix *matrix)
-{
-
-    if (matrix == NULL)
-    {
-        return;
-    }
-
-    // Speicher freigeben, falls vorhanden
-    free(matrix->buffer);
-    matrix->buffer = NULL;
-
-    // Metadaten zurücksetzen
-    matrix->rows = 0;
-    matrix->cols = 0;
-}
-
-void setMatrixAt(MatrixType value, Matrix matrix, unsigned int rowIdx, unsigned int colIdx)
-{
-    matrix.buffer[rowIdx * matrix.cols + colIdx] = value; // setzte Matrix auf den Wert value am Punkt (row col)
-}
-
-MatrixType getMatrixAt(const Matrix matrix, unsigned int rowIdx, unsigned int colIdx)
-{
-    MatrixType value = 0;
-
-    if (rowIdx < matrix.rows && colIdx < matrix.cols)
-    {
-        value = matrix.buffer[rowIdx * matrix.cols + colIdx]; // hole Wert value am Punkt (row col)
-    }
-
-    return value;
-}
-
-Matrix add(const Matrix matrix1, const Matrix matrix2)
-{
-    Matrix result = {0};
-
-    if (matrix1.rows != matrix2.rows || matrix1.cols != matrix2.cols)
-    {
-        return result;
-    }
-
-    result.rows = matrix1.rows;
-    result.cols = matrix1.cols;
-    result.buffer = malloc(result.rows * result.cols * sizeof(MatrixType));
-
-    // wenn buffer nicht allokiert werden kann dann zurücksetzen und abbrechen
-    if (result.buffer == NULL)
-    {
-        result.rows = result.cols = 0;
-
-        return result;
-    }
-
-    // Matritzenaddition
-    for (unsigned int i = 0; i < result.rows; i++)
-    {
-        for (unsigned int j = 0; j < result.cols; j++)
-        {
-            result.buffer[i * result.cols + j] = matrix1.buffer[i * matrix1.cols + j] + matrix2.buffer[i * matrix2.cols + j];
-        }
-    }
-    return result;
-}
-
-Matrix multiply(const Matrix matrix1, const Matrix matrix2)
-{
-    Matrix result = {0};
-
-    if (matrix1.rows != matrix2.rows || matrix1.cols != matrix2.cols)
-    {
-        return result;
-    }
-
-    result.rows = matrix1.rows;
-    result.cols = matrix1.cols;
-    result.buffer = malloc(result.rows * result.cols * sizeof(MatrixType));
-
-    // wenn buffer nicht allokiert werden kann dann zurücksetzen und abbrechen
-    if (result.buffer == NULL)
-    {
-        result.rows = result.cols = 0;
-
-        return result;
-    }
-
-    // Matritzenmultiplikation
-    for (unsigned int i = 0; i < result.rows; i++)
-    {
-        for (unsigned int j = 0; j < result.cols; j++)
-        {
-            MatrixType sum = 0;
-            for (unsigned int k = 0; k < matrix1.cols; k++)
-            {
-                sum += matrix1.buffer[i * matrix1.cols + k] * matrix2.buffer[k * matrix2.cols + j];
-            }
-            result.buffer[i * result.cols + j] = sum;
-        }
-    }
-
-    return result;
+#include <stdlib.h>
+#include <string.h>
+#include "matrix.h"
+
+// TODO Matrix-Funktionen implementieren
+
+Matrix createMatrix(unsigned int rows, unsigned int cols)
+{
+    Matrix m = {NULL, 0, 0};
+
+    if (rows > 0 && cols > 0)
+    {
+        m.buffer = malloc(rows * cols * sizeof(MatrixType));
+        m.rows = rows;
+        m.cols = cols;
+    }
+
+    return m;
+}
+
+void clearMatrix(Matrix *matrix)
+{
+
+    if (matrix == NULL)
+    {
+        return;
+    }
+
+    // Speicher freigeben, falls vorhanden
+    free(matrix->buffer);
+    matrix->buffer = NULL;
+
+    // Metadaten zurücksetzen
+    matrix->rows = 0;
+    matrix->cols = 0;
+}
+
+void setMatrixAt(MatrixType value, Matrix matrix, unsigned int rowIdx, unsigned int colIdx)
+{
+    matrix.buffer[rowIdx * matrix.cols + colIdx] = value; // setzte Matrix auf den Wert value am Punkt (row col)
+}
+
+MatrixType getMatrixAt(const Matrix matrix, unsigned int rowIdx, unsigned int colIdx)
+{
+    MatrixType value = 0;
+
+    if (rowIdx < matrix.rows && colIdx < matrix.cols)
+    {
+        value = matrix.buffer[rowIdx * matrix.cols + colIdx]; // hole Wert value am Punkt (row col)
+    }
+
+    return value;
+}
+
+Matrix add(const Matrix matrix1, const Matrix matrix2)
+{
+    Matrix result = {0};
+
+    int broadcast_case =
+    (matrix1.cols == 1 && matrix1.rows == matrix2.rows) ||
+    (matrix2.cols == 1 && matrix1.rows == matrix2.rows);
+
+    if (!broadcast_case && (matrix1.rows != matrix2.rows || matrix1.cols != matrix2.cols))
+    {
+        return result;
+    }
+
+
+    if (matrix1.cols == 1 && matrix1.rows == matrix2.rows) // Broadcasting
+    {
+        result = createMatrix(matrix2.rows, matrix2.cols);
+
+        for (unsigned int i = 0; i < matrix1.rows; i++)
+        {
+            for (unsigned int j = 0; j < result.cols; j++)
+            {
+                result.buffer[i * result.cols + j] = matrix1.buffer[i] + matrix2.buffer[i * matrix2.cols + j];
+            }
+        }
+        
+
+    }
+
+    else if (matrix2.cols == 1 && matrix1.rows == matrix2.rows) 
+    {
+        result = createMatrix(matrix1.rows, matrix1.cols);
+
+        for (unsigned int i = 0; i < matrix2.rows; i++)
+        {
+            for (unsigned int j = 0; j < result.cols; j++)
+            {
+                result.buffer[i * result.cols + j] = matrix1.buffer[i * matrix1.cols + j] + matrix2.buffer[i];
+            }
+        }
+    }
+
+    else
+    {
+        result = createMatrix(matrix1.rows, matrix1.cols);
+        // Elementweise Addition
+        for (unsigned int i = 0; i < result.rows; i++)
+        {
+            for (unsigned int j = 0; j < result.cols; j++)
+            {
+                result.buffer[i * result.cols + j] = matrix1.buffer[i * matrix1.cols + j] + matrix2.buffer[i * matrix2.cols + j];
+            }
+        }
+    }
+
+    
+    return result;
+}
+
+Matrix multiply(const Matrix matrix1, const Matrix matrix2)
+{
+    Matrix result = {0};
+
+    if (matrix1.cols != matrix2.rows)
+    {
+        return result;
+    }
+
+    result.rows = matrix1.rows;
+    result.cols = matrix2.cols;
+    result.buffer = malloc(result.rows * result.cols * sizeof(MatrixType));
+
+    // wenn buffer nicht allokiert werden kann dann zurücksetzen und abbrechen
+    if (result.buffer == NULL)
+    {
+        result.rows = result.cols = 0;
+
+        return result;
+    }
+
+    // Matritzenmultiplikation
+
+    for (int r = 0; r < result.rows; r++) // Zeile in Ergebnis
+    {
+        for (int m = 0; m < result.cols; m++) // Spalte in Ergebnis
+        {
+            MatrixType sum = 0;
+
+            for (int n = 0; n < matrix1.cols; n++)
+            {
+                sum += matrix1.buffer[r * matrix1.cols + n] *
+                       matrix2.buffer[n * matrix2.cols + m];
+            }
+
+            result.buffer[r * result.cols + m] = sum;
+        }
+    }
+
+    return result;
 }

matrix.h

@@ -1,26 +1,26 @@
-#ifndef MATRIX_H
-#define MATRIX_H
-
-#define UNDEFINED_MATRIX_VALUE 0
-
-typedef float MatrixType;
-
-// TODO Matrixtyp definieren
-
-typedef struct 
-{
-    unsigned int rows;   // Anzahl der Zeilen
-    unsigned int cols;   // Anzahl der Spalten
-    MatrixType *buffer;    // Zeiger auf die Matrixdaten 
-} Matrix;
-
-
-Matrix createMatrix(unsigned int rows, unsigned int cols);
-void clearMatrix(Matrix *matrix);
-void setMatrixAt(MatrixType value, Matrix matrix, unsigned int rowIdx, unsigned int colIdx);
-MatrixType getMatrixAt(const Matrix matrix, unsigned int rowIdx, unsigned int colIdx);
-Matrix add(const Matrix matrix1, const Matrix matrix2);
-Matrix multiply(const Matrix matrix1, const Matrix matrix2);
-
-
-#endif
+#ifndef MATRIX_H
+#define MATRIX_H
+
+#define UNDEFINED_MATRIX_VALUE 0
+
+typedef float MatrixType;
+
+// TODO Matrixtyp definieren
+
+typedef struct 
+{
+    MatrixType *buffer;    // Zeiger auf die Matrixdaten 
+    unsigned int rows;   // Anzahl der Zeilen
+    unsigned int cols;   // Anzahl der Spalten
+} Matrix;
+
+
+Matrix createMatrix(unsigned int rows, unsigned int cols);
+void clearMatrix(Matrix *matrix);
+void setMatrixAt(MatrixType value, Matrix matrix, unsigned int rowIdx, unsigned int colIdx);
+MatrixType getMatrixAt(const Matrix matrix, unsigned int rowIdx, unsigned int colIdx);
+Matrix add(const Matrix matrix1, const Matrix matrix2);
+Matrix multiply(const Matrix matrix1, const Matrix matrix2);
+
+
+#endif

matrixTests.c

@@ -71,6 +71,32 @@ void test_addFailsOnDifferentInputDimensions(void)
     TEST_ASSERT_EQUAL_UINT32(0, result.cols);
 }
 
+void test_addSupportsBroadcasting(void)
+{
+    MatrixType buffer1[] = {1, 2, 3, 4, 5, 6};
+    MatrixType buffer2[] = {7, 8};
+    Matrix matrix1 = {.rows=2, .cols=3, .buffer=buffer1};
+    Matrix matrix2 = {.rows=2, .cols=1, .buffer=buffer2};
+
+    Matrix result1 = add(matrix1, matrix2);
+    Matrix result2 = add(matrix2, matrix1);
+
+    float expectedResults[] = {8, 9, 10, 12, 13, 14};
+
+    TEST_ASSERT_EQUAL_UINT32(matrix1.rows, result1.rows);
+    TEST_ASSERT_EQUAL_UINT32(matrix1.cols, result1.cols);
+    TEST_ASSERT_EQUAL_UINT32(matrix1.rows, result2.rows);
+    TEST_ASSERT_EQUAL_UINT32(matrix1.cols, result2.cols);
+
+    TEST_ASSERT_EQUAL_INT(sizeof(expectedResults)/sizeof(expectedResults[0]), result1.rows * result1.cols);
+    TEST_ASSERT_EQUAL_FLOAT_ARRAY(expectedResults, result1.buffer, result1.cols * result1.rows);
+    TEST_ASSERT_EQUAL_INT(sizeof(expectedResults)/sizeof(expectedResults[0]), result2.rows * result2.cols);
+    TEST_ASSERT_EQUAL_FLOAT_ARRAY(expectedResults, result2.buffer, result2.cols * result2.rows);
+
+    free(result1.buffer);
+    free(result2.buffer);
+}
+
 void test_multiplyReturnsCorrectResults(void)
 {
     MatrixType buffer1[] = {1, 2, 3, 4, 5, 6};
@@ -159,6 +185,7 @@ int main()
     RUN_TEST(test_clearMatrixSetsMembersToNull);
     RUN_TEST(test_addReturnsCorrectResult);
     RUN_TEST(test_addFailsOnDifferentInputDimensions);
+    RUN_TEST(test_addSupportsBroadcasting);
     RUN_TEST(test_multiplyReturnsCorrectResults);
     RUN_TEST(test_multiplyFailsOnWrongInputDimensions);
     RUN_TEST(test_getMatrixAtReturnsCorrectResult);

neuralNetworkTests.c

@@ -8,7 +8,62 @@
 
 static void prepareNeuralNetworkFile(const char *path, const NeuralNetwork nn)
 {
-    // TODO
+    FILE *file = fopen(path, "wb");
+    if (!file) {
+        perror("Fehler beim Erstellen der Testdatei");
+        exit(EXIT_FAILURE);
+    }
+
+    // File header
+    const char *fileTag = "__info2_neural_network_file_format__";
+    fwrite(fileTag, strlen(fileTag), 1, file);
+
+
+    if (nn.numberOfLayers == 0)
+    {
+        unsigned int zero = 0;
+        fwrite(&zero, sizeof(unsigned int), 1, file);
+        fclose(file);
+        return;
+    }
+
+    // first layer dimension
+    unsigned int in  = nn.layers[0].weights.cols;
+    unsigned int out = nn.layers[0].weights.rows;
+
+    fwrite(&in,  sizeof(unsigned int), 1, file);
+    fwrite(&out, sizeof(unsigned int), 1, file);
+
+    // do all layers
+    for (unsigned int i = 0; i < nn.numberOfLayers; i++)
+    {
+        const Layer *L = &nn.layers[i];
+
+        // Write weights matrix
+        fwrite(L->weights.buffer,
+               sizeof(MatrixType),
+               L->weights.rows * L->weights.cols,
+               file);
+
+        // Write biases matrix
+        fwrite(L->biases.buffer,
+               sizeof(MatrixType),
+               L->biases.rows * L->biases.cols,
+               file);
+
+        // After layer i, write dimension of next layer
+        if (i + 1 < nn.numberOfLayers)
+        {
+            unsigned int nextOut = nn.layers[i+1].weights.rows;
+            fwrite(&nextOut, sizeof(unsigned int), 1, file);
+        }
+    }
+
+    // --- 5. Write terminating zero ---
+    unsigned int zero = 0;
+    fwrite(&zero, sizeof(unsigned int), 1, file);
+
+    fclose(file);
 }
 
 void test_loadModelReturnsCorrectNumberOfLayers(void)