diff --git a/Makefile b/Makefile
index 743e186..b60c9b7 100644
--- a/Makefile
+++ b/Makefile
@@ -5,7 +5,7 @@ all : myprogram
 clean :
 	rm -f *.o myprogram
 
-myprogram : randomness.o activations.o neurons.o network.o training.o main.o
+myprogram : randomness.o activations.o neurons.o network.o training.o preprocessing.o main.o
 	gcc -Wall $^ -o myprogram -lm
 
 %.o : %.c
diff --git a/main.c b/main.c
index 044ba9e..52c6c25 100644
--- a/main.c
+++ b/main.c
@@ -23,11 +23,11 @@ int main(int argc, char *argv[])
     float a=5.0;
     printf("%f\n", (float)exp((double)a));*/
     
-    int n_neurons[] = {15,120,220,120,200,25};
+    /*int n_neurons[] = {15,120,220,120,200,25};
     char *activations[] = {"relu","relu","relu","relu","relu","relu"};
     Network *net = init_network(n_neurons, 6, activations);
     print_network(net);
-    destroy_network(net);
+    destroy_network(net);*/
     
     return 0;
 }
\ No newline at end of file
diff --git a/preprocessing.c b/preprocessing.c
new file mode 100644
index 0000000..be13870
--- /dev/null
+++ b/preprocessing.c
@@ -0,0 +1,69 @@
+#include <stdio.h>
+#include <stdlib.h>
+#include "preprocessing.h"
+
+Feature *init_feature(float value)
+{
+    Feature *feature = (Feature*)malloc(sizeof(Feature));
+    feature->value = value;
+    feature->next_feature = NULL;
+
+    return feature;
+}
+
+OneHotLabel *init_onehotlabel(float value)
+{
+    OneHotLabel *hot = (OneHotLabel*)malloc(sizeof(OneHotLabel));
+    hot->value = value;
+    hot->next = NULL;
+    
+    return hot;
+}
+
+Sample *init_sample()
+{
+    Sample *sample = (Sample*)malloc(sizeof(Sample));
+    sample->first_feature = NULL;
+    sample->first_hot = NULL;
+    sample->next_sample = NULL;
+    sample->t = TRAIN;
+
+    return sample;
+}
+
+Data *init_data()
+{
+    Data *data = (Data*)malloc(sizeof(Data));
+    data->first_sample = NULL;
+    data->last_sample = NULL;
+    data->size = 0;
+
+    return data;
+}
+
+void destroy_features_from_feature(Feature *feature)
+{
+
+}
+
+void destroy_hots_from_hot(OneHotLabel *first_hot)
+{
+
+}
+
+void destroy_samples_from_sample(Sample *sample)
+{
+
+}
+
+void destroy_data(Data *data)
+{
+
+}
+
+Data *csv_to_samples(char *path_to_csv, char *features_separator, float train_percent)
+{
+    Data *data = init_data();
+    //TODO : complete
+    return data;   
+}
\ No newline at end of file
diff --git a/preprocessing.h b/preprocessing.h
new file mode 100644
index 0000000..3e3be42
--- /dev/null
+++ b/preprocessing.h
@@ -0,0 +1,51 @@
+#ifndef PREPROCESSING_H
+#define PREPROCESSING_H
+
+typedef struct feature Feature;
+struct feature
+{
+    float value;
+    Feature *next_feature;
+};
+
+typedef struct onehotlabel OneHotLabel;
+struct onehotlabel
+{
+    float value;
+    OneHotLabel *next;
+};
+
+typedef enum type Type;
+enum type
+{
+    TRAIN = 0 , TEST = 1 , VALID = 2
+};
+
+typedef struct sample Sample;
+struct sample
+{
+    Feature *first_feature;
+    OneHotLabel *first_hot;
+    Sample *next_sample;
+    Type t;
+};
+
+typedef struct data Data;
+struct data
+{
+    Sample *first_sample;
+    Sample *last_sample;
+    int size;
+};
+
+Feature *init_feature(float value);
+OneHotLabel *init_onehotlabel(float value);
+Sample *init_sample();
+Data *init_data();
+void destroy_features_from_feature(Feature *feature);
+void destroy_hots_from_hot(OneHotLabel *first_hot);
+void destroy_samples_from_sample(Sample *sample);
+void destroy_data(Data *data);
+Data *csv_to_samples(char *path_to_csv, char *features_separator, float train_percent);
+
+#endif
\ No newline at end of file