Upload 7 files
Browse files- .gitattributes +1 -0
- Main_Data.csv +0 -0
- app.py +11 -0
- eda.py +218 -0
- employee.jpg +3 -0
- model.pkl +3 -0
- prediction.py +71 -0
- requirements.txt +8 -0
.gitattributes
CHANGED
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
|
|
|
33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
36 |
+
employee.jpg filter=lfs diff=lfs merge=lfs -text
|
Main_Data.csv
ADDED
The diff for this file is too large to render.
See raw diff
|
|
app.py
ADDED
@@ -0,0 +1,11 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import eda
|
2 |
+
import prediction
|
3 |
+
import streamlit as st
|
4 |
+
|
5 |
+
|
6 |
+
page = st.sidebar.selectbox('Pilih Halaman: ', ('EDA', 'Prediction'))
|
7 |
+
|
8 |
+
if page == 'EDA':
|
9 |
+
eda.run()
|
10 |
+
else:
|
11 |
+
prediction.run()
|
eda.py
ADDED
@@ -0,0 +1,218 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import streamlit as st
|
2 |
+
import pandas as pd
|
3 |
+
import seaborn as sns
|
4 |
+
import matplotlib.pyplot as plt
|
5 |
+
import plotly.express as px
|
6 |
+
from PIL import Image
|
7 |
+
|
8 |
+
|
9 |
+
#melebarkan
|
10 |
+
st.set_page_config(
|
11 |
+
page_title='Employee Attrition Prediction',
|
12 |
+
layout='wide',
|
13 |
+
initial_sidebar_state='expanded'
|
14 |
+
|
15 |
+
)
|
16 |
+
|
17 |
+
st.markdown("""<style>.reportview-container {background: "5160549.jpg"}.sidebar .sidebar-content {background: "5160549.jpg"}</style>""",unsafe_allow_html=True)
|
18 |
+
|
19 |
+
|
20 |
+
|
21 |
+
def run():
|
22 |
+
|
23 |
+
# membuat judul
|
24 |
+
st.title('Employee Attrition Prediction')
|
25 |
+
|
26 |
+
#membuat sub header
|
27 |
+
st.subheader('Employee Attrition Prediction EDA')
|
28 |
+
|
29 |
+
|
30 |
+
|
31 |
+
# library pillow buat gambar
|
32 |
+
image = Image.open('employee.jpg')
|
33 |
+
st.markdown('---')
|
34 |
+
st.image(image, caption=' "" ')
|
35 |
+
|
36 |
+
# descripsi
|
37 |
+
st.write('### Halaman ini berisi Eksplorasi Data ')
|
38 |
+
|
39 |
+
# Membuat Garis lurus
|
40 |
+
st.markdown('---')
|
41 |
+
|
42 |
+
|
43 |
+
# Nampilin dataframe
|
44 |
+
st.write('### Employee Attrition Data')
|
45 |
+
|
46 |
+
df = pd.read_csv('Main_data.csv')
|
47 |
+
st.dataframe(df.head(5))
|
48 |
+
|
49 |
+
st.markdown('***')
|
50 |
+
#barplot
|
51 |
+
fig = plt.figure(figsize=(8,5))
|
52 |
+
|
53 |
+
|
54 |
+
###########################################
|
55 |
+
|
56 |
+
st.write('### Attrition Distribution')
|
57 |
+
|
58 |
+
# Menghitung jumlah setiap value
|
59 |
+
target_counts = df['Attrition'].value_counts()
|
60 |
+
|
61 |
+
# Membuat label untuk legenda dengan jumlah setiap value
|
62 |
+
labels = [f'Attrition {i} - {count}' for i, count in target_counts.items()]
|
63 |
+
|
64 |
+
# Membuat pie chart
|
65 |
+
fig = plt.figure(figsize=(10, 5))
|
66 |
+
plt.subplot(1, 2, 2)
|
67 |
+
target_counts.plot(kind='pie', autopct='%1.1f%%', shadow=True, labels=None, colors =['#0072C6', '#BFBFBF'])
|
68 |
+
plt.title('Employee Attrition')
|
69 |
+
|
70 |
+
# Menambahkan legenda
|
71 |
+
plt.legend(labels, loc='upper right', bbox_to_anchor=(1.3, 1))
|
72 |
+
|
73 |
+
|
74 |
+
st.pyplot(fig)
|
75 |
+
|
76 |
+
st.markdown('---')
|
77 |
+
|
78 |
+
|
79 |
+
###########################################
|
80 |
+
st.write('### Data Demografi Karyawan')
|
81 |
+
pilihan = st.selectbox('Pilih Kolom : ', ('Gender','Education','MaritalStatus','Department'))
|
82 |
+
|
83 |
+
# Melakukan pengelompokan langsung pada indeks DataFrame
|
84 |
+
attrition_data = df.groupby([df[pilihan], 'Attrition']).size().unstack(fill_value=0)
|
85 |
+
|
86 |
+
fig = plt.figure(figsize=(15, 5))
|
87 |
+
colors =['#0072C6', '#BFBFBF']
|
88 |
+
# Plot: Distribusi Attrition berdasarkan kolom yang dipilih
|
89 |
+
ax = plt.gca()
|
90 |
+
|
91 |
+
# Menyesuaikan jenis plot berdasarkan jumlah indeks attrition_data
|
92 |
+
if len(attrition_data.index) > 3:
|
93 |
+
attrition_data.plot(kind='barh', stacked=True, color=colors, ax=ax)
|
94 |
+
ax.set_xlabel('Jumlah Karyawan')
|
95 |
+
ax.set_ylabel(pilihan) # Menggunakan nama kolom yang dipilih langsung
|
96 |
+
else:
|
97 |
+
attrition_data.plot(kind='bar', stacked=True, color=colors, ax=ax)
|
98 |
+
ax.set_ylabel('Jumlah Karyawan')
|
99 |
+
ax.set_xlabel(pilihan) # Menggunakan nama kolom yang dipilih langsung
|
100 |
+
ax.set_xticklabels(attrition_data.index, rotation=0)
|
101 |
+
|
102 |
+
ax.set_title(f'Distribusi Attrition Berdasarkan {pilihan}') # Menggunakan nama kolom yang dipilih langsung
|
103 |
+
ax.legend(title='Attrition', labels=['Tidak', 'Ya'])
|
104 |
+
|
105 |
+
# Menambahkan anotasi pada plot
|
106 |
+
for container in ax.containers:
|
107 |
+
if len(attrition_data.index) > 3:
|
108 |
+
labels = [f'{int(v.get_width())}' for v in container]
|
109 |
+
else:
|
110 |
+
labels = [f'{int(v.get_height())}' for v in container]
|
111 |
+
ax.bar_label(container, labels=labels, label_type='center', padding=2)
|
112 |
+
st.pyplot(fig)
|
113 |
+
|
114 |
+
st.markdown('---')
|
115 |
+
|
116 |
+
####################################################
|
117 |
+
|
118 |
+
st.write('### Data Survey Karyawan')
|
119 |
+
pilihan = st.selectbox('Pilih Kolom : ', ('EnvironmentSatisfaction','JobSatisfaction', 'WorkLifeBalance'))
|
120 |
+
|
121 |
+
# Melakukan pengelompokan langsung pada indeks DataFrame
|
122 |
+
attrition_data = df.groupby([df[pilihan], 'Attrition']).size().unstack(fill_value=0)
|
123 |
+
|
124 |
+
fig = plt.figure(figsize=(15, 5))
|
125 |
+
colors =['#0072C6', '#BFBFBF']
|
126 |
+
# Plot: Distribusi Attrition berdasarkan kolom yang dipilih
|
127 |
+
ax = plt.gca()
|
128 |
+
|
129 |
+
# Menyesuaikan jenis plot berdasarkan jumlah indeks attrition_data
|
130 |
+
if len(attrition_data.index) > 3:
|
131 |
+
attrition_data.plot(kind='barh', stacked=True, color=colors, ax=ax)
|
132 |
+
ax.set_xlabel('Jumlah Karyawan')
|
133 |
+
ax.set_ylabel(pilihan) # Menggunakan nama kolom yang dipilih langsung
|
134 |
+
else:
|
135 |
+
attrition_data.plot(kind='bar', stacked=True, color=colors, ax=ax)
|
136 |
+
ax.set_ylabel('Jumlah Karyawan')
|
137 |
+
ax.set_xlabel(pilihan) # Menggunakan nama kolom yang dipilih langsung
|
138 |
+
ax.set_xticklabels(attrition_data.index, rotation=0)
|
139 |
+
|
140 |
+
ax.set_title(f'Distribusi Attrition Berdasarkan {pilihan}') # Menggunakan nama kolom yang dipilih langsung
|
141 |
+
ax.legend(title='Attrition', labels=['Tidak', 'Ya'])
|
142 |
+
|
143 |
+
# Menambahkan anotasi pada plot
|
144 |
+
for container in ax.containers:
|
145 |
+
if len(attrition_data.index) > 3:
|
146 |
+
labels = [f'{int(v.get_width())}' for v in container]
|
147 |
+
else:
|
148 |
+
labels = [f'{int(v.get_height())}' for v in container]
|
149 |
+
ax.bar_label(container, labels=labels, label_type='center', padding=2)
|
150 |
+
st.pyplot(fig)
|
151 |
+
|
152 |
+
st.markdown('---')
|
153 |
+
|
154 |
+
####################################################
|
155 |
+
|
156 |
+
st.write('### Data Performa Karyawan')
|
157 |
+
pilihan = st.selectbox('Pilih Kolom : ', ('JobInvolvement', 'PerformanceRating','BusinessTravel','JobLevel', 'JobRole'))
|
158 |
+
|
159 |
+
# Melakukan pengelompokan langsung pada indeks DataFrame
|
160 |
+
attrition_data = df.groupby([df[pilihan], 'Attrition']).size().unstack(fill_value=0)
|
161 |
+
|
162 |
+
fig = plt.figure(figsize=(15, 5))
|
163 |
+
colors =['#0072C6', '#BFBFBF']
|
164 |
+
# Plot: Distribusi Attrition berdasarkan kolom yang dipilih
|
165 |
+
ax = plt.gca()
|
166 |
+
|
167 |
+
# Menyesuaikan jenis plot berdasarkan jumlah indeks attrition_data
|
168 |
+
if len(attrition_data.index) > 3:
|
169 |
+
attrition_data.plot(kind='barh', stacked=True, color=colors, ax=ax)
|
170 |
+
ax.set_xlabel('Jumlah Karyawan')
|
171 |
+
ax.set_ylabel(pilihan) # Menggunakan nama kolom yang dipilih langsung
|
172 |
+
else:
|
173 |
+
attrition_data.plot(kind='bar', stacked=True, color=colors, ax=ax)
|
174 |
+
ax.set_ylabel('Jumlah Karyawan')
|
175 |
+
ax.set_xlabel(pilihan) # Menggunakan nama kolom yang dipilih langsung
|
176 |
+
ax.set_xticklabels(attrition_data.index, rotation=0)
|
177 |
+
|
178 |
+
ax.set_title(f'Distribusi Attrition Berdasarkan {pilihan}') # Menggunakan nama kolom yang dipilih langsung
|
179 |
+
ax.legend(title='Attrition', labels=['Tidak', 'Ya'])
|
180 |
+
|
181 |
+
# Menambahkan anotasi pada plot
|
182 |
+
for container in ax.containers:
|
183 |
+
if len(attrition_data.index) > 3:
|
184 |
+
labels = [f'{int(v.get_width())}' for v in container]
|
185 |
+
else:
|
186 |
+
labels = [f'{int(v.get_height())}' for v in container]
|
187 |
+
ax.bar_label(container, labels=labels, label_type='center', padding=2)
|
188 |
+
st.pyplot(fig)
|
189 |
+
|
190 |
+
st.markdown('---')
|
191 |
+
|
192 |
+
####################################################
|
193 |
+
|
194 |
+
st.write('### Data Numerical')
|
195 |
+
pilihan = st.selectbox('Pilih Kolom : ', ('Age','DistanceFromHome','MonthlyIncome', 'NumCompaniesWorked','PercentSalaryHike','TotalWorkingYears',
|
196 |
+
'YearsAtCompany','YearsSinceLastPromotion','YearsWithCurrManager'))
|
197 |
+
|
198 |
+
|
199 |
+
fig = plt.figure(figsize=(15, 5))
|
200 |
+
|
201 |
+
attrition_no = df[df['Attrition'] == 'No'][pilihan]
|
202 |
+
attrition_yes = df[df['Attrition'] == 'Yes'][pilihan]
|
203 |
+
|
204 |
+
sns.histplot(attrition_no, color=colors[0], label='No', kde=False, bins=30)
|
205 |
+
sns.histplot(attrition_yes, color=colors[1], label='Yes', kde=False, bins=30)
|
206 |
+
|
207 |
+
plt.title(f'Histogram Distribusi {pilihan} Berdasarkan Attrition')
|
208 |
+
plt.xlabel(pilihan)
|
209 |
+
plt.ylabel('Jumlah Karyawan')
|
210 |
+
plt.legend(title='Attrition')
|
211 |
+
|
212 |
+
plt.tight_layout()
|
213 |
+
|
214 |
+
st.pyplot(fig)
|
215 |
+
|
216 |
+
|
217 |
+
if __name__ == '__main__':
|
218 |
+
run()
|
employee.jpg
ADDED
Git LFS Details
|
model.pkl
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:323c9b55b492141bda434df631e0cc483a0726af23265c9ebd0829a6393cb651
|
3 |
+
size 263218
|
prediction.py
ADDED
@@ -0,0 +1,71 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import streamlit as st
|
2 |
+
import pandas as pd
|
3 |
+
import pickle
|
4 |
+
|
5 |
+
# Path to the model file
|
6 |
+
model_path = "model.pkl"
|
7 |
+
|
8 |
+
# Load the model
|
9 |
+
with open(model_path, 'rb') as f:
|
10 |
+
model = pickle.load(f)
|
11 |
+
|
12 |
+
def run():
|
13 |
+
st.title('Prediksi Pengunduran Diri Karyawan')
|
14 |
+
|
15 |
+
# Formulir untuk pengisian data
|
16 |
+
with st.form('form_employee_attrition'):
|
17 |
+
# Kolom input sesuai dengan keterangan yang Anda berikan
|
18 |
+
business_travel = st.selectbox('Business Travel', ['Travel_Rarely', 'Travel_Frequently', 'Non-Travel'])
|
19 |
+
department = st.selectbox('Department', ['Sales', 'Research & Development', 'Human Resources'])
|
20 |
+
education_field = st.selectbox('Education Field', ['Life Sciences', 'Other', 'Medical', 'Marketing', 'Technical Degree', 'Human Resources'])
|
21 |
+
job_role = st.selectbox('Job Role', ['Healthcare Representative', 'Research Scientist', 'Sales Executive', 'Human Resources', 'Research Director', 'Laboratory Technician', 'Manufacturing Director', 'Sales Representative', 'Manager'])
|
22 |
+
marital_status = st.selectbox('Marital Status', ['Married', 'Single', 'Divorced'])
|
23 |
+
training_times_last_year = st.selectbox('Training Times Last Year', [0, 1, 2, 3, 4, 5, 6])
|
24 |
+
job_involvement = st.selectbox('Job Involvement', [1, 2, 3, 4], format_func=lambda x: {1: 'Low', 2: 'Medium', 3: 'High', 4: 'Very High'}[x])
|
25 |
+
environment_satisfaction = st.selectbox('Environment Satisfaction', [1, 2, 3, 4], format_func=lambda x: {1: 'Low', 2: 'Medium', 3: 'High', 4: 'Very High'}[x])
|
26 |
+
job_satisfaction = st.selectbox('Job Satisfaction', [1, 2, 3, 4], format_func=lambda x: {1: 'Low', 2: 'Medium', 3: 'High', 4: 'Very High'}[x])
|
27 |
+
work_life_balance = st.selectbox('Work Life Balance', [1, 2, 3, 4], format_func=lambda x: {1: 'Bad', 2: 'Good', 3: 'Better', 4: 'Best'}[x])
|
28 |
+
age = st.slider('Age', min_value=18, max_value=60)
|
29 |
+
percent_salary_hike = st.slider('Percent Salary Hike', min_value=11, max_value=25)
|
30 |
+
total_working_years = st.slider('Total Working Years', min_value=0, max_value=40)
|
31 |
+
years_at_company = st.slider('Years At Company', min_value=0, max_value=40)
|
32 |
+
years_since_last_promotion = st.slider('Years Since Last Promotion', min_value=0, max_value=15)
|
33 |
+
years_with_curr_manager = st.slider('Years With Current Manager', min_value=0, max_value=17)
|
34 |
+
|
35 |
+
# Tombol untuk melakukan prediksi
|
36 |
+
submitted = st.form_submit_button('Prediksi')
|
37 |
+
|
38 |
+
# Menyusun data input menjadi DataFrame
|
39 |
+
data = {
|
40 |
+
'BusinessTravel': business_travel,
|
41 |
+
'Department': department,
|
42 |
+
'EducationField': education_field,
|
43 |
+
'JobRole': job_role,
|
44 |
+
'MaritalStatus': marital_status,
|
45 |
+
'TrainingTimesLastYear': training_times_last_year,
|
46 |
+
'JobInvolvement': job_involvement,
|
47 |
+
'EnvironmentSatisfaction': environment_satisfaction,
|
48 |
+
'JobSatisfaction': job_satisfaction,
|
49 |
+
'WorkLifeBalance': work_life_balance,
|
50 |
+
'Age': age,
|
51 |
+
'PercentSalaryHike': percent_salary_hike,
|
52 |
+
'TotalWorkingYears': total_working_years,
|
53 |
+
'YearsAtCompany': years_at_company,
|
54 |
+
'YearsSinceLastPromotion': years_since_last_promotion,
|
55 |
+
'YearsWithCurrManager': years_with_curr_manager
|
56 |
+
}
|
57 |
+
|
58 |
+
features = pd.DataFrame(data, index=[0])
|
59 |
+
|
60 |
+
# Menampilkan fitur input pengguna
|
61 |
+
st.write("## Fitur Input Pengguna")
|
62 |
+
st.write(features)
|
63 |
+
|
64 |
+
# Melakukan prediksi jika tombol prediksi ditekan
|
65 |
+
if submitted:
|
66 |
+
prediction = model.predict(features)
|
67 |
+
st.subheader('Hasil Prediksi')
|
68 |
+
st.write('Pengunduran Diri Karyawan:', 'Ya' if prediction[0] == 1 else 'Tidak')
|
69 |
+
|
70 |
+
if __name__ == '__main__':
|
71 |
+
run()
|
requirements.txt
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
streamlit
|
2 |
+
pandas
|
3 |
+
seaborn
|
4 |
+
matplotlib
|
5 |
+
numpy
|
6 |
+
scikit-learn==1.2.2
|
7 |
+
Pillow
|
8 |
+
plotly
|