Part 15 — Manipulasi Data III: Group & Summarise

Fundamental Series — Part 15 of 20

Split-Apply-Combine — pisahkan data per grup, hitung sesuatu, gabungkan hasilnya. Ini adalah operasi paling fundamental di data analysis.

Setup Data

import pandas as pd
import polars as pl

data = {
    "nama": ["Andi","Budi","Citra","Dina","Eko","Fani","Gita","Hadi"],
    "kota": ["Jakarta","Bandung","Jakarta","Surabaya","Bandung","Jakarta","Surabaya","Bandung"],
    "dept": ["IT","IT","HR","HR","IT","HR","IT","HR"],
    "gaji": [5000,7000,6000,8000,4500,5500,7500,6500]
}

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

GroupBy + Agg — Pandas

# Satu kolom, satu fungsi
df_pd.groupby("kota")["gaji"].mean()

# Multiple fungsi
df_pd.groupby("kota")["gaji"].agg(["mean", "std", "count"])

# Multiple kolom dan fungsi
(df_pd
    .groupby("kota")
    .agg(
        mean_gaji=("gaji", "mean"),
        n=("nama", "count")
    )
    .reset_index()
)

# Multiple grouping
df_pd.groupby(["kota", "dept"])["gaji"].mean().reset_index()

GroupBy + Agg — Polars

df_pl.group_by("kota").agg(
    pl.col("gaji").mean().alias("mean_gaji"),
    pl.col("gaji").std().alias("sd_gaji"),
    pl.col("nama").count().alias("n")
)

# Multiple grouping
df_pl.group_by(["kota", "dept"]).agg(
    pl.col("gaji").mean().alias("mean_gaji")
)

Count / Value Counts

# Pandas
df_pd["kota"].value_counts()
df_pd.groupby("kota").size().reset_index(name="n")

# Polars
df_pl.group_by("kota").len()
df_pl["kota"].value_counts()

Transform (Window per Grup)

# Pandas — tambah kolom berdasarkan grup (tanpa collapse)
df_pd["mean_gaji_kota"] = df_pd.groupby("kota")["gaji"].transform("mean")
df_pd["persen"] = df_pd["gaji"] / df_pd.groupby("kota")["gaji"].transform("sum") * 100

# Polars — over()
df_pl.with_columns(
    pl.col("gaji").mean().over("kota").alias("mean_gaji_kota"),
    (pl.col("gaji") / pl.col("gaji").sum().over("kota") * 100).alias("persen")
)

Fungsi Agregasi Umum

Pandas	Polars	Keterangan
`"mean"`	`.mean()`	Rata-rata
`"median"`	`.median()`	Median
`"std"`	`.std()`	Standar deviasi
`"sum"`	`.sum()`	Total
`"min"` / `"max"`	`.min()` / `.max()`	Min / Max
`"count"`	`.count()`	Jumlah non-null
`"nunique"`	`.n_unique()`	Jumlah unik
`"first"` / `"last"`	`.first()` / `.last()`	Pertama / terakhir

Latihan

Latihan 15.1

# 1. Hitung mean gaji per departemen (pandas + polars)
# 2. Hitung mean gaji per kota + dept
# 3. Kota mana total gajinya tertinggi?

Latihan 15.2

# 1. Tambah kolom: gaji sebagai persen total gaji per dept
# 2. Ranking gaji per dept (tertinggi = 1)
# 3. Filter rank == 1 di setiap dept

Ringkasan

Operasi	Pandas	Polars
Group + Agg	`.groupby().agg()`	`.group_by().agg()`
Count	`.value_counts()`	`.value_counts()`
Window/Transform	`.groupby().transform()`	`.over()`

Sebelumnya: Part 14 — Mutate & Arrange Selanjutnya: Part 16 — Reshape & Merge

--- title: "Part 15 — Manipulasi Data III: Group & Summarise" description: "Agregasi data per grup di Python: groupby, agg, transform, value_counts dengan pandas dan polars." date: "2026-02-26" categories: [Fundamental, DataManipulation] series: "Fundamental Programming" order: 15 --- ## Fundamental Series — Part 15 of 20 **Split-Apply-Combine** — pisahkan data per grup, hitung sesuatu, gabungkan hasilnya. Ini adalah operasi paling fundamental di data analysis. --- ## Setup Data ```python import pandas as pd import polars as pl data = { "nama": ["Andi","Budi","Citra","Dina","Eko","Fani","Gita","Hadi"], "kota": ["Jakarta","Bandung","Jakarta","Surabaya","Bandung","Jakarta","Surabaya","Bandung"], "dept": ["IT","IT","HR","HR","IT","HR","IT","HR"], "gaji": [5000,7000,6000,8000,4500,5500,7500,6500] } df_pd = pd.DataFrame(data) df_pl = pl.DataFrame(data) ``` --- ## GroupBy + Agg — Pandas ```python # Satu kolom, satu fungsi df_pd.groupby("kota")["gaji"].mean() # Multiple fungsi df_pd.groupby("kota")["gaji"].agg(["mean", "std", "count"]) # Multiple kolom dan fungsi (df_pd .groupby("kota") .agg( mean_gaji=("gaji", "mean"), n=("nama", "count") ) .reset_index() ) # Multiple grouping df_pd.groupby(["kota", "dept"])["gaji"].mean().reset_index() ``` --- ## GroupBy + Agg — Polars ```python df_pl.group_by("kota").agg( pl.col("gaji").mean().alias("mean_gaji"), pl.col("gaji").std().alias("sd_gaji"), pl.col("nama").count().alias("n") ) # Multiple grouping df_pl.group_by(["kota", "dept"]).agg( pl.col("gaji").mean().alias("mean_gaji") ) ``` --- ## Count / Value Counts ```python # Pandas df_pd["kota"].value_counts() df_pd.groupby("kota").size().reset_index(name="n") # Polars df_pl.group_by("kota").len() df_pl["kota"].value_counts() ``` --- ## Transform (Window per Grup) ```python # Pandas — tambah kolom berdasarkan grup (tanpa collapse) df_pd["mean_gaji_kota"] = df_pd.groupby("kota")["gaji"].transform("mean") df_pd["persen"] = df_pd["gaji"] / df_pd.groupby("kota")["gaji"].transform("sum") * 100 # Polars — over() df_pl.with_columns( pl.col("gaji").mean().over("kota").alias("mean_gaji_kota"), (pl.col("gaji") / pl.col("gaji").sum().over("kota") * 100).alias("persen") ) ``` --- ## Fungsi Agregasi Umum | Pandas | Polars | Keterangan | |--------|--------|-----------| | `"mean"` | `.mean()` | Rata-rata | | `"median"` | `.median()` | Median | | `"std"` | `.std()` | Standar deviasi | | `"sum"` | `.sum()` | Total | | `"min"` / `"max"` | `.min()` / `.max()` | Min / Max | | `"count"` | `.count()` | Jumlah non-null | | `"nunique"` | `.n_unique()` | Jumlah unik | | `"first"` / `"last"` | `.first()` / `.last()` | Pertama / terakhir | --- ## Latihan ::: {.callout-caution} ## Latihan 15.1 ```python # 1. Hitung mean gaji per departemen (pandas + polars) # 2. Hitung mean gaji per kota + dept # 3. Kota mana total gajinya tertinggi? ``` ::: ::: {.callout-caution} ## Latihan 15.2 ```python # 1. Tambah kolom: gaji sebagai persen total gaji per dept # 2. Ranking gaji per dept (tertinggi = 1) # 3. Filter rank == 1 di setiap dept ``` ::: --- ## Ringkasan | Operasi | Pandas | Polars | |---------|--------|--------| | Group + Agg | `.groupby().agg()` | `.group_by().agg()` | | Count | `.value_counts()` | `.value_counts()` | | Window/Transform | `.groupby().transform()` | `.over()` | --- *Sebelumnya: [Part 14 — Mutate & Arrange](../2026_02_26_14_mutate_arrange/index.qmd)* *Selanjutnya: [Part 16 — Reshape & Merge](../2026_02_26_16_reshape_merge/index.qmd)*