Scripts project updates

.gitignore

@@ -1,8 +1,8 @@
 # project data
-data
-saved_models
-threshold_map
-models_info
+data/*
+saved_models/*
+threshold_map/*
+models_info/*
 
 fichiersSVD_light/*/*/*.csv
 fichiersSVD_light/*_min_max_values

README.md

@@ -7,7 +7,7 @@ pip install -r requirements.txt
 ```
 
 Generate all needed data for each metrics
-```
+```bash
 python generate_all_data.py --metric all
 ```
 
@@ -18,7 +18,7 @@ For noise detection, many metrics are available :
 - low_bits_4
 
 You can also specify metric you want to compute :
-```
+```bash
 python generate_all_data.py --metric mscn
 ```
 
@@ -63,7 +63,7 @@ Parameters explained :
 
 This is an example of how to train a model
 
-```python
+```bash
 python models/xxxxx.py --data 'data/xxxxx.train' --output 'model_file_to_save'
 ```
 
@@ -71,7 +71,7 @@ python models/xxxxx.py --data 'data/xxxxx.train' --output 'model_file_to_save'
 
 Now we have a model trained, we can use it with an image as input :
 
-```python
+```bash
 python metrics_predictions/predict_noisy_image_svd_lab.py --image path/to/image.png --interval "x,x" --model saved_models/xxxxxx.joblib --mode 'svdn'
 ```
 
@@ -81,7 +81,7 @@ The model will return only 0 or 1 :
 
 You can also use other specific metric
 
-```python
+```bash
 python metrics_predictions/predict_noisy_image_svd_mscn.py --image path/to/image.png --interval "x,x" --model saved_models/xxxxxx.joblib --mode 'svdn'
 ```
 
@@ -94,7 +94,7 @@ All SVD metrics you developed need :
 
 Now we have a model trained, we can use it with an image as input :
 
-```python
+```bash
 python prediction_scene.py --data path/to/xxxx.csv --model saved_model/xxxx.joblib --output xxxxx --scene xxxx
 ```
 **Remark** : *scene* parameter expected need to be the correct name of the Scene.
@@ -121,7 +121,7 @@ Parameters list :
 
 Main objective of this project is to predict as well as a human the noise perception on a photo realistic image. Human threshold is available from training data. So a script was developed to give the predicted treshold from model and compare predicted treshold from the expected one.
 
-```python
+```bash
 python predict_seuil_expe.py --interval "x,x" --model 'saved_models/xxxx.joblib' --mode ["svd", "svdn", "svdne"] --metric ['lab', 'mscn', ...] --limit_detection xx
 ```
 
@@ -141,7 +141,7 @@ The content will be divised into two parts :
 
 The previous script need to already have ran to obtain and display treshold maps on this markdown file.
 
-```python
+```bash
 python save_model_result_in_md.py --interval "xx,xx" --model saved_models/xxxx.joblib --mode ["svd", "svdn", "svdne"] --metric ['lab', 'mscn'] 
 ```
 

generateAndTrain_maxwell.sh

@@ -14,7 +14,7 @@ if [ -z "$2" ]
     exit 1
 fi
 
-
+result_filename="models_info/models_comparisons.csv"
 VECTOR_SIZE=200
 size=$1
 metric=$2
@@ -25,42 +25,49 @@ scenes="A, D, G, H"
 half=$(($size/2))
 start=-$half
 for counter in {0..4}; do
-end=$(($start+$size))
+    end=$(($start+$size))
 
-if [ "$end" -gt "$VECTOR_SIZE" ]; then
-    start=$(($VECTOR_SIZE-$size))
-    end=$(($VECTOR_SIZE))
-fi
+    if [ "$end" -gt "$VECTOR_SIZE" ]; then
+        start=$(($VECTOR_SIZE-$size))
+        end=$(($VECTOR_SIZE))
+    fi
 
-if [ "$start" -lt "0" ]; then
-    start=$((0))
-    end=$(($size))
-fi
+    if [ "$start" -lt "0" ]; then
+        start=$((0))
+        end=$(($size))
+    fi
 
-for nb_zones in {6,8,10,12,16}; do
+    for nb_zones in {6,8,10,12,16}; do
 
-    echo $start $end
+        echo $start $end
 
-    for mode in {"svd","svdn","svdne"}; do
-        for model in {"svm_model","ensemble_model","ensemble_model_v2"}; do
+        for mode in {"svd","svdn","svdne"}; do
+            for model in {"svm_model","ensemble_model","ensemble_model_v2"}; do
 
-            FILENAME="data/data_maxwell_N${size}_B${start}_E${end}_nb_zones_${nb_zones}_${metric}_${mode}"
-            MODEL_NAME="${model}_N${size}_B${start}_E${end}_nb_zones_${nb_zones}_${metric}_${mode}"
+                FILENAME="data/data_maxwell_N${size}_B${start}_E${end}_nb_zones_${nb_zones}_${metric}_${mode}"
+                MODEL_NAME="${model}_N${size}_B${start}_E${end}_nb_zones_${nb_zones}_${metric}_${mode}"
 
-            echo $FILENAME
-            python generate_data_model_random.py --output ${FILENAME} --interval "${start},${end}" --kind ${mode} --metric ${metric} --scenes "${scenes}" --nb_zones "${nb_zones}" --percent 1 --sep ';' --rowindex '0'
-            python models/${model}_train.py --data ${FILENAME}.train --output ${MODEL_NAME}
-           
-            python predict_seuil_expe_maxwell.py --interval "${start},${end}" --model "saved_models/${MODEL_NAME}.joblib" --mode "${mode}" --metric ${metric} --limit_detection '2'
-            python save_model_result_in_md_maxwell.py --interval "${start},${end}" --model "saved_models/${MODEL_NAME}.joblib" --mode "${mode}" --metric ${metric}
+                echo $FILENAME
+                
+                # only compute if necessary (perhaps server will fall.. Just in case)
+                if grep -q "${MODEL_NAME}" "${result_filename}"; then
+
+                    echo "${MODEL_NAME} results already generated..."
+                else
+                    python generate_data_model_random.py --output ${FILENAME} --interval "${start},${end}" --kind ${mode} --metric ${metric} --scenes "${scenes}" --nb_zones "${nb_zones}" --percent 1 --sep ';' --rowindex '0'
+                    python models/${model}_train.py --data ${FILENAME}.train --output ${MODEL_NAME}
+                
+                    python predict_seuil_expe_maxwell.py --interval "${start},${end}" --model "saved_models/${MODEL_NAME}.joblib" --mode "${mode}" --metric ${metric} --limit_detection '2'
+                    python save_model_result_in_md_maxwell.py --interval "${start},${end}" --model "saved_models/${MODEL_NAME}.joblib" --mode "${mode}" --metric ${metric}
+                fi
+            done
         done
     done
-done
 
-if [ "$counter" -eq "0" ]; then
-    start=$(($start+50-$half))
-else 
-    start=$(($start+50))
-fi
+    if [ "$counter" -eq "0" ]; then
+        start=$(($start+50-$half))
+    else 
+        start=$(($start+50))
+    fi
 
 done

models/ensemble_model_train.py

@@ -14,8 +14,9 @@ from sklearn.metrics import accuracy_score
 
 import sys, os, getopt
 
+saved_models_folder = 'saved_models'
 current_dirpath = os.getcwd()
-output_model_folder = os.path.join(current_dirpath, 'saved_models')
+output_model_folder = os.path.join(current_dirpath, saved_models_folder)
 
 def get_best_model(X_train, y_train):
     Cs = [0.001, 0.01, 0.1, 1, 10]
@@ -100,6 +101,10 @@ def main():
     y_pred = ensemble_model.predict(X_test)
     print("**Test :** " + str(accuracy_score(y_test, y_pred)))
 
+    # create path if not exists
+    if not os.path.exists(saved_models_folder):
+        os.makedirs(saved_models_folder)
+
     joblib.dump(ensemble_model, output_model_folder + '/' + p_output + '.joblib') 
 
 if __name__== "__main__":

models/ensemble_model_train_v2.py

@@ -16,8 +16,10 @@ from sklearn.metrics import accuracy_score
 
 import sys, os, getopt
 
+
+saved_models_folder = 'saved_models'
 current_dirpath = os.getcwd()
-output_model_folder = os.path.join(current_dirpath, 'saved_models')
+output_model_folder = os.path.join(current_dirpath, saved_models_folder)
 
 def get_best_model(X_train, y_train):
     Cs = [0.001, 0.01, 0.1, 1, 10, 20, 30]
@@ -107,6 +109,10 @@ def main():
     y_pred = ensemble_model.predict(X_test)
     print("**Test :** " + str(accuracy_score(y_test, y_pred)))
 
+    # create path if not exists
+    if not os.path.exists(saved_models_folder):
+        os.makedirs(saved_models_folder)
+
     joblib.dump(ensemble_model, output_model_folder + '/' +  p_output + '.joblib') 
 
 if __name__== "__main__":

models/svm_model_train.py

@@ -13,8 +13,9 @@ from sklearn.metrics import accuracy_score
 
 import sys, os, getopt
 
+saved_models_folder = 'saved_models'
 current_dirpath = os.getcwd()
-output_model_folder = os.path.join(current_dirpath, 'saved_models')
+output_model_folder = os.path.join(current_dirpath, saved_models_folder)
 
 def get_best_model(X_train, y_train):
     
@@ -84,6 +85,10 @@ def main():
     y_pred = svm_model.predict(X_test)
     print("**Test :** " + str(accuracy_score(y_test, y_pred)))
 
+    # create path if not exists
+    if not os.path.exists(saved_models_folder):
+        os.makedirs(saved_models_folder)
+        
     joblib.dump(svm_model, output_model_folder + '/' + p_output + '.joblib') 
 
 if __name__== "__main__":

predict_seuil_expe_maxwell.py

@@ -167,7 +167,7 @@ def main():
             # end of scene => display of results
 
             # construct path using model name for saving threshold map folder
-            model_treshold_path = os.path.join(threshold_map_folder, p_model_file.split('/')[-1].replace('.joblib', '_'))
+            model_treshold_path = os.path.join(threshold_map_folder, p_model_file.split('/')[-1].replace('.joblib', ''))
             
             # create threshold model path if necessary
             if not os.path.exists(model_treshold_path):

runAll_maxwell.sh

@@ -12,6 +12,6 @@ echo 'model_name; vector_size; start; end; nb_zones; metric; mode; train; test;
 for size in {"4","8","16","26","32","40"}; do
 
     for metric in {"lab","mscn","low_bits_4","low_bits_2"}; do
-        bash generateAndTrain_maxwell.sh ${size} ${metric}
+        bash generateAndTrain_maxwell.sh ${size} ${metric} &
     done
 done