DataSetMaker start

README.md

@@ -26,6 +26,6 @@ $ (venv) python3 app/visualisation.py
 
 ### Measurements visualization of dataset FD001
 
-[<img src="app/assets/eda.png"/>](app/assets/image.png) 
+[<img src="app/assets/eda.png" width=100% height=600/>](app/assets/eda.png) 
 
 

app/callbacks.py

@@ -11,14 +11,13 @@
 def get_callbacks(app):
     @app.callback(
             Output('interval', 'n_intervals'),
-            Output({'type':'graph','id':ALL}, 'data'),
             Input('toggle-button','n_clicks'),
             State('interval','n_intervals'),
             State('interval','max_intervals'),
     )
     def start_interval_again(n,current_n,max_n):
         if current_n == max_n:
-            return 0, [[] for _ in range(len(dash.callback_context.outputs_list[1]))]
+            return 0#, [[] for _ in range(len(dash.callback_context.outputs_list[1]))]
         else:
             raise dash.exceptions.PreventUpdate
 
@@ -107,9 +106,6 @@ def disable_dropdowns(n):
             disabled =bool(n%2)
             return disabled,disabled,disabled
         raise dash.exceptions.PreventUpdate
-
-
-
 
     @app.callback(
         Output({'type':'graph','id':ALL}, 'extendData'),
@@ -118,11 +114,11 @@ def disable_dropdowns(n):
         prevent_initial_call=True
     )
     def update_operational_setting_graph(n_intervals,data):
-        n_of_graphs = len(dash.callback_context.outputs_list)
         output_graphs = (dash.callback_context.outputs_list)
         if n_intervals==0:
             raise dash.exceptions.PreventUpdate
         s = []
+
         result = [entry for entry in data if entry.get('time') == n_intervals][0]
         for x in output_graphs:
             current_value = x['id']['id'].split('graph-')[1]
@@ -135,7 +131,7 @@ def update_operational_setting_graph(n_intervals,data):
     prevent_initial_call=True
     )
     def take_data(value):
-        return make_dataframe('./data/'+value)
+        return make_dataframe(value)
 
     @app.callback(
         Output('unit-slider', 'min'),

app/layout.py

@@ -8,18 +8,18 @@
 N = 6
 df = pd.DataFrame()
 
-directory_path = './data/'
+directory_path = './data/raw/'
 prefix_list = ['train_', 'test_']
 
 layout = html.Div(
     children=[
-        html.H2("Prediction of Remaining Useful Life of JET engine",style={"textAlign":"center"}),
+        html.H2("Fault Detection of a JET Engine",style={"textAlign":"center"}),
         dbc.Row(
             style={'margin':'1%'},
             children=[
             dbc.Col(width=2,className="dash-bootstrap",children=[
                     html.Label("FD choice"),
-                    dcc.Dropdown(options = [{'label':x.replace('.txt', ''),'value':x} for x in get_files_with_prefix(directory_path, prefix_list)], id='fd-choice-dropdown'),
+                    dcc.Dropdown(options = [{'label':x.replace('.txt', '').replace(directory_path,''),'value':x} for x in get_files_with_prefix(directory_path, prefix_list)], id='fd-choice-dropdown'),
             ]),
                 dbc.Col(width=4,children=[
                     html.Label("Unit choice"),
@@ -55,7 +55,7 @@
             id="rul-and-operational",
             style={'margin':'1%'},
             children=[dbc.Col(width=6, children=[
-                    html.P("RUL PREDICTION"),
+                    html.P("FAULT DETECTION"),
                     dcc.Graph(
                         id = {'type':'graph','id':f'graph-RUL'},
                         config={'displayModeBar': False},

app/visualisation.py

@@ -17,4 +17,4 @@
 
 
 if __name__ == '__main__':
-    app.run_server(port=1129, debug=False)
+    app.run_server(port=1129, debug=True)

modelling/dataset_creation.py

@@ -0,0 +1,51 @@
+import pandas as pd
+
+class DataSetMaker:
+    def __init__(self,train_set:bool, path1:str, path2:str=None):
+        self.train_set = train_set
+        self.path1 = path1
+        self.path2 = path2
+
+    def make_dataframe(self) -> bool:
+        if self.train_set:
+            data = pd.read_csv(self.path1)
+
+
+if __name__ == '__main__':
+    s = DataSetMaker(True,"a").make_dataframe()
+    print(s)
+
+
+
+def make_dataframe(text_file_path):
+    column_names = ['unit_number','time','operational_setting_1','operational_setting_2','operational_setting_3']
+    N =24 
+    column_names += [f'sensor_measurement_{i}' for i in range(1,N)]
+    df = pd.read_csv(text_file_path, sep=" ",header=None,names=column_names)
+    # deleting empty columns
+    df.dropna(how='all', axis=1, inplace=True) 
+    #delete duplicates
+    df.drop_duplicates(inplace=True)
+    grouped_by = df.groupby(by='unit_number')
+    max_time = grouped_by['time'].max()
+    df = df.merge(max_time.to_frame(name='max_time'), left_on='unit_number', right_index=True)
+    df['RUL'] = df['max_time'] - df['time']
+    df['fault_detected'] = df.groupby('unit_number')['RUL'].transform(detect_fault)
+    return df.to_dict('records')
+
+def detect_fault(series):
+    warning = 0.7
+    fault = 0.1
+    total_cycles = series.max()  # Maximum cycle for the unit
+    threshold_warning = warning * total_cycles
+    threshold_fault = fault * total_cycles
+
+    def fault_detection(rul):
+        if rul > threshold_warning:
+            return 0  # Everything is ok
+        elif rul > threshold_fault:
+            return 1  # Warning
+        else:
+            return 2  # Fault detected
+
+    return series.apply(fault_detection)