Python untuk Data Science: Panduan Memulai

# List — kumpulan data berurutan
suhu_harian = [28.5, 31.2, 29.8, 33.1, 27.6, 30.0, 32.5]
print(f"Suhu rata-rata: {sum(suhu_harian) / len(suhu_harian):.1f}°C")
print(f"Suhu tertinggi: {max(suhu_harian)}°C")
print(f"Suhu terendah: {min(suhu_harian)}°C")

# Slicing — mengambil subset data
tiga_hari_pertama = suhu_harian[:3]      # [28.5, 31.2, 29.8]
tiga_hari_terakhir = suhu_harian[-3:]    # [30.0, 32.5] → tidak benar, 3 terakhir

# Dictionary — data berformat key-value
mahasiswa = {
    "nama": "Budi Santoso",
    "nim": "2024001",
    "ipk": 3.75,
    "semester": 4,
    "mata_kuliah": ["Algoritma", "Statistika", "Basis Data"]
}

print(f"Nama: {mahasiswa['nama']}")
print(f"IPK: {mahasiswa['ipk']}")
print(f"Jumlah MK: {len(mahasiswa['mata_kuliah'])}")

# List Comprehension — membuat list baru secara ringkas
suhu_celsius = [28.5, 31.2, 29.8, 33.1, 27.6]

# Konversi ke Fahrenheit
suhu_fahrenheit = [c * 9/5 + 32 for c in suhu_celsius]
print(suhu_fahrenheit)  # [83.3, 88.16, 85.64, 91.58, 81.68]

# Filter: ambil suhu di atas 30°C
suhu_panas = [c for c in suhu_celsius if c > 30]
print(suhu_panas)  # [31.2, 33.1]

# Dictionary Comprehension
suhu_dict = {f"Hari {i+1}": s for i, s in enumerate(suhu_celsius)}
print(suhu_dict)
# {'Hari 1': 28.5, 'Hari 2': 31.2, 'Hari 3': 29.8, ...}

# Fungsi biasa
def hitung_statistik(data):
    """Menghitung mean, min, max dari data numerik."""
    n = len(data)
    mean = sum(data) / n
    return {
        "count": n,
        "mean": round(mean, 2),
        "min": min(data),
        "max": max(data),
        "range": max(data) - min(data)
    }

nilai_ujian = [85, 92, 78, 95, 88, 76, 90, 83]
stats = hitung_statistik(nilai_ujian)
print(stats)
# {'count': 8, 'mean': 85.88, 'min': 76, 'max': 95, 'range': 19}

# Lambda — fungsi satu baris yang ringkas
kali_dua = lambda x: x * 2
print(kali_dua(5))  # 10

# Lambda sangat berguna dengan fungsi sorting
data_produk = [
    {"nama": "Laptop", "harga": 12000000},
    {"nama": "Mouse", "harga": 150000},
    {"nama": "Keyboard", "harga": 850000},
    {"nama": "Monitor", "harga": 4500000},
]

# Sort berdasarkan harga (termurah ke termahal)
data_sorted = sorted(data_produk, key=lambda x: x["harga"])
for p in data_sorted:
    print(f"  {p['nama']}: Rp {p['harga']:,}")

import csv

# Membaca file CSV dengan modul bawaan Python
with open('data_penjualan.csv', 'r', encoding='utf-8') as f:
    reader = csv.DictReader(f)
    data = list(reader)

print(f"Jumlah baris: {len(data)}")
print(f"Kolom: {list(data[0].keys())}")

# Menghitung total penjualan
total = sum(float(row['harga']) * int(row['jumlah']) for row in data)
print(f"Total penjualan: Rp {total:,.0f}")

# Buat virtual environment khusus Data Science
python3 -m venv ds-env
source ds-env/bin/activate  # Linux/Mac
# ds-env\Scripts\activate   # Windows

# Instalasi Jupyter dan library Data Science
pip install jupyter numpy pandas matplotlib seaborn scikit-learn

# Jalankan Jupyter Notebook
jupyter notebook

# Alternatif: JupyterLab (interface lebih modern)
pip install jupyterlab
jupyter lab

import numpy as np

# Membuat array dari list Python
arr1d = np.array([1, 2, 3, 4, 5])
print(f"1D Array: {arr1d}")
print(f"Tipe: {type(arr1d)}")         # <class 'numpy.ndarray'>
print(f"Shape: {arr1d.shape}")        # (5,)
print(f"Dimensi: {arr1d.ndim}")       # 1
print(f"Tipe data: {arr1d.dtype}")    # int64

# Membuat 2D array (matrix)
arr2d = np.array([
    [1, 2, 3],
    [4, 5, 6],
    [7, 8, 9]
])
print(f"2D Array:\n{arr2d}")
print(f"Shape: {arr2d.shape}")        # (3, 3) — 3 baris, 3 kolom

# Array khusus
zeros = np.zeros((3, 4))              # 3x4 matrix berisi 0
ones = np.ones((2, 3))                # 2x3 matrix berisi 1
eye = np.eye(4)                       # 4x4 identity matrix
range_arr = np.arange(0, 20, 2)       # [0, 2, 4, 6, 8, 10, 12, 14, 16, 18]
linspace = np.linspace(0, 1, 5)       # [0, 0.25, 0.5, 0.75, 1.0]

# Array random
np.random.seed(42)                    # Untuk reproducibility
random_uniform = np.random.rand(3, 3)         # Uniform [0, 1)
random_normal = np.random.randn(3, 3)          # Normal (mean=0, std=1)
random_int = np.random.randint(1, 100, size=5) # Random int [1, 100)
print(f"Random integers: {random_int}")

import numpy as np

a = np.array([10, 20, 30, 40, 50])
b = np.array([5, 4, 3, 2, 1])

# Operasi element-wise (tanpa loop!)
print(f"Penjumlahan:  {a + b}")        # [15 24 33 42 51]
print(f"Pengurangan:  {a - b}")        # [ 5 16 27 38 49]
print(f"Perkalian:    {a * b}")        # [50 80 90 80 50]
print(f"Pembagian:    {a / b}")        # [2.  5. 10. 20. 50.]
print(f"Pangkat:      {a ** 2}")       # [100 400 900 1600 2500]
print(f"Akar kuadrat: {np.sqrt(a)}")   # [3.16 4.47 5.48 6.32 7.07]

# Statistik
data = np.random.normal(loc=170, scale=10, size=1000)  # 1000 data tinggi badan
print(f"\nStatistik Tinggi Badan:")
print(f"  Mean:   {np.mean(data):.2f} cm")
print(f"  Median: {np.median(data):.2f} cm")
print(f"  Std:    {np.std(data):.2f} cm")
print(f"  Min:    {np.min(data):.2f} cm")
print(f"  Max:    {np.max(data):.2f} cm")
print(f"  P25:    {np.percentile(data, 25):.2f} cm")
print(f"  P75:    {np.percentile(data, 75):.2f} cm")

# Indexing & Slicing (mirip list tapi lebih powerful)
matrix = np.array([[1, 2, 3], [4, 5, 6], [7, 8, 9]])
print(f"Baris 0, Kolom 1: {matrix[0, 1]}")    # 2
print(f"Baris terakhir: {matrix[-1, :]}")       # [7 8 9]
print(f"Kolom tengah: {matrix[:, 1]}")          # [2 5 8]

# Boolean indexing (filter data)
tinggi = np.array([165, 180, 155, 190, 172, 168, 175])
mask = tinggi > 170  # [False, True, False, True, True, False, True]
tinggi_tinggi = tinggi[mask]
print(f"Tinggi > 170: {tinggi_tinggi}")  # [180 190 172 175]

# Reshape
arr = np.arange(12)
matrix_3x4 = arr.reshape(3, 4)
print(f"3x4 Matrix:\n{matrix_3x4}")

import pandas as pd

# Membuat DataFrame dari dictionary
data_penjualan = {
    'tanggal': ['2026-01-01', '2026-01-01', '2026-01-02', '2026-01-02',
                '2026-01-03', '2026-01-03', '2026-01-04', '2026-01-04'],
    'produk': ['Laptop', 'Mouse', 'Laptop', 'Keyboard',
               'Mouse', 'Monitor', 'Laptop', 'Keyboard'],
    'harga': [12000000, 150000, 12000000, 850000,
              150000, 4500000, 12000000, 850000],
    'jumlah': [2, 10, 1, 5, 15, 3, 4, 8]
}

df = pd.DataFrame(data_penjualan)
print(df)
#      tanggal    produk     harga  jumlah
# 0  2026-01-01   Laptop  12000000       2
# 1  2026-01-01    Mouse    150000      10
# 2  2026-01-02   Laptop  12000000       1
# ...

# Membaca file CSV (paling umum)
# df = pd.read_csv('data.csv')
# df = pd.read_excel('data.xlsx')
# df = pd.read_json('data.json')

# Info dasar DataFrame
print(f"Shape: {df.shape}")             # (8, 4) — 8 baris, 4 kolom
print(f"Kolom: {list(df.columns)}")     # ['tanggal', 'produk', 'harga', 'jumlah']
print(f"\nInfo:")
print(df.info())
print(f"\nStatistik Deskriptif:")
print(df.describe())

# Seleksi kolom
produk = df['produk']            # Series
subset = df[['produk', 'harga']] # DataFrame (2 kolom)

# Seleksi baris dengan loc (berdasarkan label) dan iloc (berdasarkan index)
baris_0 = df.iloc[0]              # Baris pertama
baris_2_4 = df.iloc[2:5]          # Baris index 2, 3, 4

# Filtering — sangat powerful!
laptop_only = df[df['produk'] == 'Laptop']
print(f"Laptop rows:\n{laptop_only}")

# Multiple filter (& = AND, | = OR)
filter_laptop_mahal = df[(df['produk'] == 'Laptop') & (df['harga'] > 10000000)]
filter_mouse_atau_keyboard = df[df['produk'].isin(['Mouse', 'Keyboard'])]

# Filter dengan string contains
nama_produk_nama_L = df[df['produk'].str.startswith('L')]

# Sorting
df_sorted = df.sort_values('harga', ascending=False)
print(f"\nDiurutkan dari termahal:\n{df_sorted.head()}")

# Tambah kolom total_transaksi
df['total_transaksi'] = df['harga'] * df['jumlah']

# Group by produk
per_produk = df.groupby('produk').agg({
    'jumlah': 'sum',              # Total jumlah terjual
    'total_transaksi': 'sum',     # Total pendapatan
    'harga': 'mean'               # Rata-rata harga
}).round(0)

print("Ringkasan per Produk:")
print(per_produk.sort_values('total_transaksi', ascending=False))

# Group by tanggal
per_tanggal = df.groupby('tanggal')['total_transaksi'].sum()
print(f"\nPendapatan per Tanggal:")
print(per_tanggal)

# Pivot table
pivot = df.pivot_table(
    values='jumlah',
    index='tanggal',
    columns='produk',
    aggfunc='sum',
    fill_value=0
)
print(f"\nPivot Table:")
print(pivot)

import pandas as pd
import numpy as np

# Contoh DataFrame dengan missing values
df = pd.DataFrame({
    'nama': ['Alice', 'Budi', 'Charlie', 'Dewi', 'Eka', 'Farhan'],
    'usia': [25, np.nan, 30, 22, np.nan, 28],
    'kota': ['Jakarta', 'Bandung', np.nan, 'Surabaya', 'Bandung', np.nan],
    'gaji': [8000000, 6500000, 9000000, np.nan, 7000000, 7500000]
})

print("Data awal:")
print(df)
print(f"\nMissing values per kolom:")
print(df.isnull().sum())
# nama    0
# usia    2
# kota    2
# gaji    1

# Strategi 1: Hapus baris yang memiliki missing values
df_dropped = df.dropna()
print(f"\nSetelah dropna (baris lengkap): {len(df_dropped)}")

# Strategi 2: Isi missing values
df_filled = df.copy()
df_filled['usia'] = df_filled['usia'].fillna(df_filled['usia'].median())
df_filled['gaji'] = df_filled['gaji'].fillna(df_filled['gaji'].mean())
df_filled['kota'] = df_filled['kota'].fillna('Tidak Diketahui')

print(f"\nSetelah fillna:")
print(df_filled)

# Strategi 3: Forward fill (untuk data time series)
# df['nilai'] = df['nilai'].ffill()  # Isi dengan nilai sebelumnya

# Strategi 4: Interpolate (untuk data numerik berurutan)
# df['suhu'] = df['suhu'].interpolate(method='linear')

# Deteksi duplikasi
df_dup = pd.DataFrame({
    'nama': ['Alice', 'Budi', 'Alice', 'Dewi', 'Budi'],
    'kota': ['Jakarta', 'Bandung', 'Jakarta', 'Surabaya', 'Bandung']
})

print(f"Jumlah duplikat: {df_dup.duplicated().sum()}")

# Hapus duplikasi
df_clean = df_dup.drop_duplicates()
print(f"Setelah hapus duplikat: {len(df_clean)} baris")

# Konversi tipe data
df_tipe = pd.DataFrame({
    'tanggal': ['2026-01-15', '2026-02-20', '2026-03-10'],
    'harga_str': ['1500000', '2500000', '3500000'],
    'diskon_str': ['10%', '15%', '20%']
})

# Konversi string ke datetime
df_tipe['tanggal'] = pd.to_datetime(df_tipe['tanggal'])
print(f"Tipe tanggal: {df_tipe['tanggal'].dtype}")  # datetime64[ns]

# Konversi string ke numeric
df_tipe['harga'] = pd.to_numeric(df_tipe['harga_str'])

# Konversi persen string ke float
df_tipe['diskon'] = df_tipe['diskon_str'].str.replace('%', '').astype(float) / 100

print(df_tipe)
# Ekstrak komponen tanggal
df_tipe['bulan'] = df_tipe['tanggal'].dt.month
df_tipe['hari_nama'] = df_tipe['tanggal'].dt.day_name()

import numpy as np
import pandas as pd

# Contoh dataset: nilai ujian 30 mahasiswa
np.random.seed(42)
nilai = np.random.normal(loc=75, scale=12, size=30).round(1)
nilai = np.clip(nilai, 0, 100)  # Batasi antara 0-100

print("=== STATISTIK DESKRIPTIF ===")
print(f"Data: {nilai}")
print(f"\nUkuran Tendensi Sentral:")
print(f"  Mean (Rata-rata): {np.mean(nilai):.2f}")
print(f"  Median:           {np.median(nilai):.2f}")
print(f"  Modus:            {pd.Series(nilai).mode().values[0]:.2f}")

print(f"\nUkuran Penyebaran:")
print(f"  Range:            {np.ptp(nilai):.2f}")
print(f"  Variance:         {np.var(nilai, ddof=1):.2f}")
print(f"  Std Deviation:    {np.std(nilai, ddof=1):.2f}")
print(f"  Q1 (25%):         {np.percentile(nilai, 25):.2f}")
print(f"  Q3 (75%):         {np.percentile(nilai, 75):.2f}")
print(f"  IQR:              {np.percentile(nilai, 75) - np.percentile(nilai, 25):.2f}")

# Deskripsi lengkap dengan Pandas
df_nilai = pd.DataFrame({'nilai': nilai})
print(f"\n=== DESKRIPSI PANDAS ===")
print(df_nilai.describe())

# Deteksi outlier dengan IQR method
Q1 = np.percentile(nilai, 25)
Q3 = np.percentile(nilai, 75)
IQR = Q3 - Q1
lower_bound = Q1 - 1.5 * IQR
upper_bound = Q3 + 1.5 * IQR

outliers = nilai[(nilai < lower_bound) | (nilai > upper_bound)]
print(f"\nOutlier detection (IQR method):")
print(f"  Lower bound: {lower_bound:.2f}")
print(f"  Upper bound: {upper_bound:.2f}")
print(f"  Outliers: {outliers}")

# Korelasi antar variabel
np.random.seed(42)
df_corr = pd.DataFrame({
    'jam_belajar': np.random.uniform(1, 10, 50).round(1),
    'kehadiran': np.random.uniform(60, 100, 50).round(1),
})
# Buat nilai_ujian berkorelasi positif dengan jam_belajar
df_corr['nilai_ujian'] = (
    df_corr['jam_belajar'] * 5
    + df_corr['kehadiran'] * 0.3
    + np.random.normal(0, 5, 50)
).round(1)

# Matriks korelasi
corr_matrix = df_corr.corr()
print("Matriks Korelasi:")
print(corr_matrix.round(3))

# Interpretasi korelasi:
#  0.0 - 0.3: Lemah
#  0.3 - 0.7: Sedang
#  0.7 - 1.0: Kuat
# Nilai negatif = korelasi terbalik

import matplotlib.pyplot as plt
import numpy as np

# ===== LINE CHART: Tren Penjualan Bulanan =====
bulan = ['Jan', 'Feb', 'Mar', 'Apr', 'Mei', 'Jun',
         'Jul', 'Agu', 'Sep', 'Okt', 'Nov', 'Des']
penjualan_2025 = [120, 135, 148, 130, 155, 170, 160, 180, 195, 175, 200, 210]
penjualan_2026 = [140, 155, 168, 150, 175, 190, 180, 200, 215, 195, 220, 230]

plt.figure(figsize=(12, 5))
plt.plot(bulan, penjualan_2025, 'b-o', label='2025', linewidth=2, markersize=6)
plt.plot(bulan, penjualan_2026, 'r-s', label='2026', linewidth=2, markersize=6)
plt.title('Tren Penjualan Bulanan (dalam juta Rp)', fontsize=14, fontweight='bold')
plt.xlabel('Bulan')
plt.ylabel('Penjualan (Juta Rp)')
plt.legend()
plt.grid(True, alpha=0.3)
plt.tight_layout()
plt.savefig('tren_penjualan.png', dpi=150)
plt.show()

# ===== BAR CHART: Perbandingan per Kategori =====
kategori = ['Elektronik', 'Fashion', 'Makanan', 'Kesehatan', 'Olahraga']
pendapatan = [450, 380, 520, 290, 310]
warna = ['#FF6B6B', '#4ECDC4', '#45B7D1', '#96CEB4', '#FFEAA7']

plt.figure(figsize=(10, 5))
bars = plt.bar(kategori, pendapatan, color=warna, edgecolor='white', linewidth=1.5)

# Tambah label di atas bar
for bar, val in zip(bars, pendapatan):
    plt.text(bar.get_x() + bar.get_width()/2, bar.get_height() + 5,
             f'{val}jt', ha='center', fontsize=10, fontweight='bold')

plt.title('Pendapatan per Kategori (Juta Rp)', fontsize=14, fontweight='bold')
plt.ylabel('Pendapatan (Juta Rp)')
plt.ylim(0, max(pendapatan) * 1.15)
plt.grid(axis='y', alpha=0.3)
plt.tight_layout()
plt.savefig('pendapatan_kategori.png', dpi=150)
plt.show()

# ===== HISTOGRAM: Distribusi Usia Karyawan =====
np.random.seed(42)
usia = np.random.normal(loc=35, scale=8, size=200).astype(int)
usia = np.clip(usia, 22, 60)

plt.figure(figsize=(10, 5))
plt.hist(usia, bins=15, color='#45B7D1', edgecolor='white', alpha=0.85)
plt.axvline(np.mean(usia), color='red', linestyle='--', linewidth=2,
            label=f'Mean: {np.mean(usia):.1f} tahun')
plt.axvline(np.median(usia), color='green', linestyle='--', linewidth=2,
            label=f'Median: {np.median(usia):.1f} tahun')
plt.title('Distribusi Usia Karyawan', fontsize=14, fontweight='bold')
plt.xlabel('Usia (tahun)')
plt.ylabel('Jumlah Karyawan')
plt.legend()
plt.grid(axis='y', alpha=0.3)
plt.tight_layout()
plt.savefig('distribusi_usia.png', dpi=150)
plt.show()

# ===== SCATTER PLOT: Hubungan Jam Belajar vs Nilai =====
np.random.seed(42)
jam_belajar = np.random.uniform(1, 10, 50)
nilai_ujian = 40 + jam_belajar * 5 + np.random.normal(0, 8, 50)
nilai_ujian = np.clip(nilai_ujian, 0, 100)

plt.figure(figsize=(10, 6))
scatter = plt.scatter(jam_belajar, nilai_ujian, c=nilai_ujian,
                      cmap='RdYlGn', s=100, edgecolor='gray', alpha=0.8)
plt.colorbar(scatter, label='Nilai Ujian')

# Tambah trend line
z = np.polyfit(jam_belajar, nilai_ujian, 1)
p = np.poly1d(z)
x_line = np.linspace(1, 10, 100)
plt.plot(x_line, p(x_line), 'r--', linewidth=2, label=f'Trend: y={z[0]:.1f}x+{z[1]:.1f}')

plt.title('Hubungan Jam Belajar vs Nilai Ujian', fontsize=14, fontweight='bold')
plt.xlabel('Jam Belajar per Hari')
plt.ylabel('Nilai Ujian')
plt.legend()
plt.grid(True, alpha=0.3)
plt.tight_layout()
plt.savefig('scatter_belajar_nilai.png', dpi=150)
plt.show()

# Membuat dashboard dengan 4 chart
fig, axes = plt.subplots(2, 2, figsize=(14, 10))

# Chart 1: Line
axes[0, 0].plot(bulan, penjualan_2026, 'r-o', linewidth=2)
axes[0, 0].set_title('Tren Penjualan 2026')
axes[0, 0].set_ylabel('Juta Rp')
axes[0, 0].grid(True, alpha=0.3)

# Chart 2: Bar
axes[0, 1].bar(kategori, pendapatan, color=warna)
axes[0, 1].set_title('Pendapatan per Kategori')
axes[0, 1].set_ylabel('Juta Rp')

# Chart 3: Histogram
axes[1, 0].hist(usia, bins=12, color='#45B7D1', edgecolor='white')
axes[1, 0].set_title('Distribusi Usia')
axes[1, 0].set_xlabel('Usia')

# Chart 4: Pie
axes[1, 1].pie(pendapatan, labels=kategori, autopct='%1.1f%%',
               colors=warna, startangle=90)
axes[1, 1].set_title('Proporsi Pendapatan')

fig.suptitle('Dashboard Analitik', fontsize=16, fontweight='bold')
plt.tight_layout()
plt.savefig('dashboard.png', dpi=150, bbox_inches='tight')
plt.show()