Tipe Data I:
Klasifikasi & Karakteristik

import pandas as pd
import numpy as np

# Dataset mahasiswa dengan berbagai tipe data
data = {
    'nim':           ['2021001', '2021002', '2021003', '2021004', '2021005'],
    'nama':          ['Andi', 'Budi', 'Citra', 'Dewi', 'Eko'],
    'jenis_kelamin': ['L', 'L', 'P', 'P', 'L'],          # nominal
    'grade':         ['B', 'A', 'A', 'C', 'B'],             # ordinal
    'jumlah_mk':     [6, 7, 6, 5, 7],                      # diskrit
    'ipk':           [3.45, 3.78, 3.92, 2.85, 3.60],       # kontinu (rasio)
    'tgl_masuk':     ['2021-09-01']*5,                       # tanggal (string → datetime)
    'aktif':         [True, True, True, False, True]         # boolean
}

df = pd.DataFrame(data)
print("=== INFO AWAL (DTYPE DEFAULT) ===")
print(df.dtypes)
print(f"\nMemori digunakan: {df.memory_usage(deep=True).sum():,} bytes")

# ── 1. Nominal → category ────────────────────────────
df['jenis_kelamin'] = df['jenis_kelamin'].astype('category')

# ── 2. Ordinal → category dengan urutan ─────────────
grade_order = pd.CategoricalDtype(
    categories=['D', 'C', 'B', 'A'],
    ordered=True   # D < C < B < A
)
df['grade'] = df['grade'].astype(grade_order)

# ── 3. String → datetime64 ──────────────────────────
df['tgl_masuk'] = pd.to_datetime(df['tgl_masuk'])

# ── 4. Cek hasil konversi ───────────────────────────
print("=== SETELAH KONVERSI ===")
print(df.dtypes)
print(f"\nMemori setelah: {df.memory_usage(deep=True).sum():,} bytes")

# ── 5. Manfaat urutan ordinal ───────────────────────
print("\n=== MANFAAT ORDINAL TERURUT ===")
print("Apakah grade 'A' > 'B'? :", df['grade'][1] > df['grade'][0])
print("Mahasiswa dengan grade ≥ B:")
print(df[df['grade'] >= 'B'][['nama', 'grade', 'ipk']])

import matplotlib.pyplot as plt

# Dataset lebih besar untuk visualisasi
np.random.seed(42)
n = 100
df_besar = pd.DataFrame({
    'jk'   : np.random.choice(['L','P'], n),
    'grade': np.random.choice(['A','B','C','D'], n, p=[.25,.40,.25,.10]),
    'mk'   : np.random.randint(4, 9, n),
    'ipk'  : np.random.normal(3.2, 0.4, n).clip(0, 4)
})

fig, axes = plt.subplots(1, 3, figsize=(14, 4))

# ① Nominal → Bar chart (frekuensi)
df_besar['jk'].value_counts().plot(kind='bar', ax=axes[0],
    color=['#3B82F6','#C084FC'], edgecolor='white')
axes[0].set_title('Jenis Kelamin (Nominal)\n→ Bar Chart Frekuensi')

# ② Ordinal → Bar chart terurut
grade_counts = df_besar['grade'].value_counts().reindex(['A','B','C','D'])
grade_counts.plot(kind='bar', ax=axes[1],
    color=['#22D3EE','#3B82F6','#F59E0B','#F87171'], edgecolor='white')
axes[1].set_title('Grade (Ordinal)\n→ Bar Chart Berurutan A→D')

# ③ Kontinu → Histogram
axes[2].hist(df_besar['ipk'], bins=15, color='#4ADE80',
             edgecolor='white', alpha=0.85)
axes[2].set_title('IPK (Kontinu/Rasio)\n→ Histogram Distribusi')
axes[2].axvline(df_besar['ipk'].mean(), color='red',
               linestyle='--', label=f"Mean={df_besar['ipk'].mean():.2f}")
axes[2].legend()

plt.tight_layout()
plt.show()