job_salary.py

# -*- coding: utf-8 -*-
"""JobSalary.ipynb

Automatically generated by Colaboratory.

Original file is located at
    https://colab.research.google.com/drive/16gJawevsBMslRMByU4TvTaB7emBhCKkh

<a href="https://githubtocolab.com/giswqs/geemap/blob/master/examples/notebooks/35_geemap_colab.ipynb" target="_parent"><img src="https://colab.research.google.com/assets/colab-badge.svg" alt="Open in Colab"/></a>
"""

from google.colab import files
uploaded = files.upload()

# Import Library
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
from sklearn.model_selection import train_test_split
from sklearn.model_selection import GridSearchCV
from sklearn.metrics import roc_auc_score

df_train = pd.read_csv('train.csv')
df_test = pd.read_csv('test.csv')
df = [df_train, df_test]

df_train.head()

df_train = df_train.replace({
    '?': np.nan,
    '<=7jt': 0,
    '>7jt': 1
    })
df_test = df_test.replace('?', np.nan)
df = [df_train, df_test]

df_train.head()

df_train.info()
print("\n\n")
df_test.info()

df_train.describe()

# Analyze Data
# 'Kelas Pekerja' and 'Gaji'
df_train.groupby('Kelas Pekerja')[['Gaji']].mean()

# 'Pendidikan' and 'Gaji'
df_train.groupby('Pendidikan')[['Gaji']].mean()

# 'Status' and 'Gaji'
df_train.groupby('Status Perkawinan')[['Gaji']].mean()

# 'Pekerjaan' and 'Gaji'
df_train.groupby('Pekerjaan')[['Gaji']].mean()

# 'Jenis Kelamin' and 'Gaji'
df_train.groupby('Jenis Kelamin')[['Gaji']].mean()

# Visualizing Data
# 'Umur' and 'Gaji'
grid = sns.FacetGrid(df_train, col='Gaji')
grid.map(sns.kdeplot, 'Umur')

# 'Jenis Kelamin', 'Status Perkawinan', and 'Gaji'
grid = sns.FacetGrid(df_train, col='Jenis Kelamin', height=4)
grid.map(sns.barplot, 'Gaji', 'Status Perkawinan', ci=None)
grid.add_legend()

grid = sns.FacetGrid(df_train, col='Jenis Kelamin')
grid.map(sns.barplot, 'Gaji', 'Berat Akhir', ci=None)
grid.add_legend();

# Data Processing
# Handle Missing Value in 'Kelas Pekerja'
df_train['Kelas Pekerja'].value_counts()

for dataset in df:
  dataset['Kelas Pekerja'] = dataset['Kelas Pekerja'].fillna('Wiraswasta')

df_train.head()

# Handle Missing Value in 'Pekerjaan'
df_train['Pekerjaan'].value_counts()

for dataset in df:
  dataset['Pekerjaan'] = dataset['Pekerjaan'].fillna('Spesialis')

df_train.head()

# Convert 'Kelas Pekerja'
conv_kerja = {
    'Pemerintah Negara': 0,
    'Pemerintah Provinsi': 1,
    'Pemerintah Lokal': 2,
    'Pekerja Bebas Perusahaan': 3,
    'Pekerja Bebas Bukan Perusahan': 4,
    'Wiraswasta': 5,
    'Tidak Pernah Bekerja': 6,
    'Tanpa di Bayar': 7
}

for dataset in df:
  dataset['Kelas Pekerja'] = dataset['Kelas Pekerja'].map(conv_kerja)

df_train.head()

# Convert 'Pendidikan'
for dataset in df:
  dataset['Pendidikan'] = dataset['Pendidikan'].replace(['1st-4th', '5th-6th'], 'SD')
  dataset['Pendidikan'] = dataset['Pendidikan'].replace(['7th-8th', '9th'], 'SMP')
  dataset['Pendidikan'] = dataset['Pendidikan'].replace(['10th', '11th', '12th'], 'SMA')
  dataset['Pendidikan'] = dataset['Pendidikan'].replace(['Sarjana', 'D4'], 'D4/Sarjana')
  dataset['Pendidikan'] = dataset['Pendidikan'].replace('Sekolah Professional', 'Pendidikan Tinggi')

df_train.groupby('Pendidikan')[['Gaji']].mean()

conv_didik = {
    'SD': 0,
    'SMP': 1,
    'SMA': 2,
    'D3': 3,
    'D4/Sarjana': 4,
    'Master': 5,
    'Doktor': 6,
    'Pendidikan Tinggi': 7
}

for dataset in df:
  dataset['Pendidikan'] = dataset['Pendidikan'].map(conv_didik)

df_train.head()

df_train.groupby('Status Perkawinan')[['Gaji']].mean()

# Convert 'Status Perkawinan'
for dataset in df:
  dataset['Status Perkawinan'] = dataset['Status Perkawinan'].replace(['Belum Pernah Menikah', 'Berpisah', 'Cerai', 'Janda'], 'Tidak Menikah')
  dataset['Status Perkawinan'] = dataset['Status Perkawinan'].replace('Menikah LDR', 'Menikah')

df_train.groupby('Status Perkawinan')[['Gaji']].mean()

for dataset in df:
  dataset['Status Perkawinan'] = dataset['Status Perkawinan'].map({
      'Tidak Menikah': 0,
      'Menikah': 1
  })

df_train.head()

# Convert 'Pekerjaan'
conv_kerja = {
    'Asisten Rumah Tangga': 0,
    'Ekesekutif Managerial': 1,
    'Mesin Inspeksi': 2,
    'Pembersih': 3,
    'Pemuka Agama': 4,
    'Penjaga': 5,
    'Perbaikan Kerajinan': 6,
    'Petani': 7,
    'Sales': 8,
    'Servis Lainnya': 9,
    'Spesialis': 10,
    'Supir': 11,
    'Tech-support': 12,
    'Tentara': 13,
}

for dataset in df:
  dataset['Pekerjaan'] = dataset['Pekerjaan'].map(conv_kerja)

df_train.head()

# Convert 'Jenis Kelamin'
for dataset in df:
  dataset['Jenis Kelamin'] = dataset['Jenis Kelamin'].map({
      'Perempuan': 0,
      'Laki2': 1
  })

df_train.head()

# There are too many zero values in 'Kerugian Capital' and 'Keuntungan Capital'
# There are no correlation 'Jumlah Tahun Pendidikan' with 'Gaji'
df_train = df_train.drop(['Kerugian Capital', 'Keuntungan Kapital', 'Jmlh Tahun Pendidikan'], axis=1)
df_test = df_test.drop(['Kerugian Capital', 'Keuntungan Kapital', 'Jmlh Tahun Pendidikan'], axis=1)
df = [df_train, df_test]

df_train.info()

df_train.head()

# Modeling
X = df_train.drop(['id', 'Gaji'], axis=1)
y = df_train['Gaji']
# X_test = df_test.drop(['id'], axis=1).copy()

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, stratify=y)

"""

---



---

"""

from sklearn.ensemble import RandomForestClassifier

model = RandomForestClassifier(n_estimators=151, criterion='entropy', random_state=0)
model.fit(X_train, y_train)

# Predicting with Test Data
y_test_pred = model.predict(X_test)
y_test_pred

# Scoring
eval = roc_auc_score(y_test, y_test_pred)
eval

"""

---



---



---

"""

from sklearn.svm import SVC

model = SVC()
model.fit(X_train, y_train)

# Predicting with Test Data
y_test_pred = model.predict(X_test)
y_test_pred

# Scoring
eval = roc_auc_score(y_test, y_test_pred)
eval

"""

---



---



---

"""

from sklearn.neighbors import KNeighborsClassifier

model = KNeighborsClassifier()
model.fit(X_train, y_train)

# Predicting with Test Data
y_test_pred = model.predict(X_test)
y_test_pred

# Scoring
eval = roc_auc_score(y_test, y_test_pred)
eval

"""

---



---



---

"""

from sklearn.tree import DecisionTreeClassifier

model = DecisionTreeClassifier()
model.fit(X_train, y_train)

# Predicting with Test Data
y_test_pred = model.predict(X_test)
y_test_pred

# Scoring
eval = roc_auc_score(y_test, y_test_pred)
eval

"""

---



---



---

"""

# Predicting with New Data
X_pred = df_test.drop(['id'], axis=1).copy()

y_pred = model.predict(X_pred)
y_pred

result = pd.DataFrame({
    'id': df_test['id'],
    'Gaji': y_pred
})

result

result.to_csv('submission_gaji1.csv', index=False)