Part 14 — Manipulasi Data II: Mutate & Arrange

Fundamental Series — Part 14 of 20

Setelah filter dan select, operasi berikutnya yang paling sering: membuat kolom baru dan mengurutkan data.

---

Setup Data

import pandas as pd
import polars as pl
import numpy as np

data = {
    "nama": ["Andi", "Budi", "Citra", "Dina", "Eko"],
    "kota": ["Jakarta", "Bandung", "Jakarta", "Surabaya", "Bandung"],
    "umur": [25, 30, 28, 35, 22],
    "gaji": [5000, 7000, 6000, 8000, 4500]
}

df_pd = pd.DataFrame(data)
df_pl = pl.DataFrame(data)

---

Tambah Kolom — Pandas

# Direct assignment
df_pd["gaji_ribu"] = df_pd["gaji"] * 1000

# assign() — chainable
(df_pd
    .assign(
        gaji_ribu=lambda x: x["gaji"] * 1000,
        umur_next=lambda x: x["umur"] + 1
    )
)

---

Tambah Kolom — Polars

df_pl.with_columns(
    (pl.col("gaji") * 1000).alias("gaji_ribu"),
    (pl.col("umur") + 1).alias("umur_next")
)

# Mengubah kolom yang ada
df_pl.with_columns(
    (pl.col("gaji") * 1000).alias("gaji")
)

---

Conditional Column

Pandas — np.where dan np.select

# 2 kondisi
df_pd["senior"] = np.where(df_pd["umur"] >= 30, "Ya", "Tidak")

# Multiple kondisi
conditions = [
    df_pd["umur"] < 25,
    df_pd["umur"] < 30,
    df_pd["umur"] < 35
]
choices = ["Muda", "Dewasa Awal", "Dewasa"]
df_pd["kategori"] = np.select(conditions, choices, default="Senior")

Polars — when/then/otherwise

df_pl.with_columns(
    pl.when(pl.col("umur") < 25).then(pl.lit("Muda"))
      .when(pl.col("umur") < 30).then(pl.lit("Dewasa Awal"))
      .when(pl.col("umur") < 35).then(pl.lit("Dewasa"))
      .otherwise(pl.lit("Senior"))
      .alias("kategori")
)

---

Sort — Pandas

# Ascending
df_pd.sort_values("gaji")

# Descending
df_pd.sort_values("gaji", ascending=False)

# Multiple kolom
df_pd.sort_values(["kota", "gaji"], ascending=[True, False])

---

Sort — Polars

df_pl.sort("gaji")
df_pl.sort("gaji", descending=True)
df_pl.sort(["kota", "gaji"], descending=[False, True])

---

Ranking

# Pandas
df_pd["rank_gaji"] = df_pd["gaji"].rank(ascending=False)

# Polars
df_pl.with_columns(
    pl.col("gaji").rank(descending=True).alias("rank_gaji")
)

---

Latihan

BahayaLatihan 14.1

# 1. Tambah kolom gaji_juta = gaji / 1000
# 2. Tambah kolom kategori: "tinggi" jika gaji > 6000, else "rendah"
# 3. Urutkan dari gaji tertinggi ke terendah
# Kerjakan dengan pandas DAN polars

BahayaLatihan 14.2

# 1. Tambah kolom z_gaji = (gaji - mean) / std
# 2. Ranking gaji (tertinggi = 1)
# 3. Pilih hanya nama, gaji, z_gaji, rank

---

Ringkasan

Operasi	Pandas	Polars
Tambah kolom	.assign() / df["col"]=	.with_columns()
Conditional	np.where() / np.select()	pl.when().then()
Sort	.sort_values()	.sort()
Rank	.rank()	pl.col().rank()

---

Sebelumnya: Part 13 — Filter & Select Selanjutnya: Part 15 — Manipulasi Data III: Group & Summarise