Merge pull request #20 from HazelJJJ/main

update project structure and add reflection

Procfile

@@ -1 +1 @@
-web: gunicorn src.app:server
+web: gunicorn src.app:server --pythonpath=src

app.py

@@ -0,0 +1,115 @@
+import dash
+import dash_html_components as html
+import dash_core_components as dcc
+from dash.dependencies import Input, Output
+import altair as alt
+import pandas as pd
+import dash_bootstrap_components as dbc
+
+#updated data path
+df = pd.read_csv(r"data/Processed/HR_employee_Attrition_editted.csv")
+###Convert variables to categoriccal and reordering by label.
+df['EnvironmentSatisfaction'] = df['EnvironmentSatisfaction'].astype('category').cat.rename_categories(["Bad", "Good", "Better", "Best"])
+df['WorkLifeBalance']=df['WorkLifeBalance'].astype('category').cat.rename_categories(["Low", "Medium", "High", "Very High"])
+df["Department"]=df["Department"].astype('category')
+df["BusinessTravel"]=df["BusinessTravel"].astype('category')
+df['BusinessTravel'] = df['BusinessTravel'].cat.rename_categories(["No Travel", "Travel Frequently", "Travel Rarely"])
+#df['BusinessTravel'] = df['BusinessTravel'].cat.reorder_categories(["Travel Rarely", "Travel Frequently", "No Travel"], ordered=True)
+
+# Setup app and layout/frontend
+app = dash.Dash(external_stylesheets=[dbc.themes.BOOTSTRAP])
+server = app.server
+app.layout = dbc.Container([
+    html.H1('Key Factors for Employee Attrition'),
+    dbc.Row([
+        dbc.Col([
+            'Department',
+            dcc.Dropdown(
+                id='depart-widget',
+                value='Sales',
+                options=[
+                    {'label': col, 'value': col} for col in list(set(df.Department.tolist()))],
+                placeholder='Select a department'),
+
+            # 'Job Role',
+            # dcc.Dropdown(
+            #     id='job-widget',
+            #     value='Sales Representative',
+            #     options=[
+            #         {'label': col, 'value': col} for col in set(df.JobRole.tolist())],
+            #     placeholder='Select a job role'),
+
+            'Gender',
+            dcc.Dropdown(
+                id='gender-widget',
+                value='Female',  # REQUIRED to show the plot on the first page load
+                options=[
+                    {'label': "Female", 'value': "Female"},
+                    {'label': "Male", 'value': "Male"}],
+                placeholder='Select gender'
+                ),
+
+            'Age',
+            dcc.RangeSlider(
+                id='age_slider',
+                min=df['Age'].min(),
+                max=df['Age'].max(),
+                value=[18, 60],
+                step=1,
+                marks={30: '30', 40: '40', 50:'50'},
+                tooltip={'always_visible': True, 'placement': 'bottomLeft'})],
+            md=4,),
+
+        dbc.Col(
+            html.Iframe(
+                id='scatter',
+                style={'border-width': '0', 'width': '200%', 'height': '800px'}),
+                )])])  
+            #style={"height": "1000vh"}
+
+# Set up callbacks/backend
+@app.callback(
+    Output('scatter', 'srcDoc'),
+    Input('depart-widget', 'value'),
+    #Input('job-widget', 'value'),
+    Input('gender-widget', 'value'),
+    Input('age_slider', 'value'))
+
+def plot_altair(depart,gender, age=18):
+    chart_income = alt.Chart(
+        df[(df['Department']==depart) & (df['Gender']==gender)&(df['Age']>=age[0])&(df['Age']<=age[1])], 
+        title='Monthly Income Distribution'
+        ).mark_boxplot(size = 50).encode(
+        x=alt.X('MonthlyIncome:Q', scale=alt.Scale(zero=False), axis=alt.Axis(grid=False)),
+        y=alt.Y('Attrition',  axis=alt.Axis(grid=False)),
+        color='Attrition').properties(height=200, width=250)
+
+    chart_worklife = alt.Chart(
+        df[(df['Department']==depart) &(df['Gender']==gender)&(df['Age']>=age[0])&(df['Age']<=age[1])], 
+        title='Work Life Balance').mark_bar().encode(
+        y=alt.Y('WorkLifeBalance:O', title='', scale=alt.Scale(domain=["Low", "Medium", "High", "Very High"])),
+        x=alt.X('count()', stack = 'normalize', axis=alt.Axis(format='%'), title = 'Proportion'),
+        color = 'Attrition'
+    ).properties(height=200, width=250)
+
+    chart_travel = alt.Chart(
+        df[(df['Department']==depart) &(df['Gender']==gender)&(df['Age']>=age[0])&(df['Age']<=age[1])], 
+        title='Business Travel Frequency').mark_bar().encode(
+        y=alt.Y("BusinessTravel", title="", scale=alt.Scale(domain=["No Travel", "Travel Rarely", "Travel Frequently"]),
+        x=alt.X('count()', stack="normalize", axis=alt.Axis(format='%'), title='Proportion'),
+        color = "Attrition").properties(height=200, width=250))
+
+    chart_environment = alt.Chart(
+        df[(df['Department']==depart) & (df['Gender']==gender)&(df['Age']>=age[0])&(df['Age']<=age[1])], 
+        title='Environment Satisfaction').mark_bar().encode(
+        y=alt.Y('EnvironmentSatisfaction', title='', scale=alt.Scale(domain=["Bad", "Good", "Better", "Best"])),
+        x=alt.X('count()', stack = 'normalize', axis=alt.Axis(format='%'), title = 'Proportion'),
+        color='Attrition').properties(height=200, width=250)
+
+    chart = (chart_income&chart_travel)|(chart_worklife&chart_environment)
+    return chart.to_html()
+
+
+
+if __name__ == '__main__':
+    app.run_server(debug=True)