# Python Programing # Pandas Cheatsheet # Cheatsheet 1 In **pandas**, there are a few quick + reliable ways to check the **data type of columns**, depending on what you need. --- #### 1️⃣ Check *all* column types (most common)
`df.dtypes`
**Output example**
`age int64height float64name objectactive booldate datetime64[ns]dtype: object`
--- #### 2️⃣ Detailed overview (very useful for debugging)
`df.info()`
Shows: - column name - non-null count - dtype - memory usage --- #### 3️⃣ Check type of a *single column*
`df["age"].dtype`
or
`df.age.dtype`
--- #### 4️⃣ Check only numeric / non-numeric columns
`df.select_dtypes(include="number").columns`
`df.select_dtypes(exclude="number").columns`
--- #### 5️⃣ Check if a column is a specific type
`import pandas as pdpd.api.types.is_numeric_dtype(df["age"])pd.api.types.is_string_dtype(df["name"])pd.api.types.is_datetime64_any_dtype(df["date"])pd.api.types.is_bool_dtype(df["active"])`
--- #### 6️⃣ Count how many columns per dtype
`df.dtypes.value_counts()`
--- #### 7️⃣ Detect mixed or problematic types (object columns)
`df.select_dtypes(include="object").apply( lambda s: s.map(type).value_counts())`
Useful when: - CSV imported badly - numbers stored as strings - mixed types in one column
TaskCommand
See all dtypes`df.dtypes`
Full summary`df.info()`
Single column`df[col].dtype`
Only numeric`select_dtypes()`
Type check`pd.api.types.is_*`
# Cheatsheet 2 Oke, ini **cheat-sheet Python (pandas)** buat **subset, grouping, dan filtering** — dari **paling dasar sampai yang sering dipakai di data real**. Contohnya aku sesuaikan dengan data kamu (provinsi, kab, gender, dll). --- # 1️⃣ SUBSET (ambil sebagian data) ## a) Subset kolom
`df[["provinsi", "kab", "gender_ketua_kelompok"]]`
## b) Subset baris (index-based)
`df.iloc[:10] # 10 baris pertamadf.iloc[5:20] # slice`
`df.loc[100:200] # berdasarkan index label`
## c) Subset baris + kondisi
`df[df["provinsi"] == "JAWA BARAT"]`
--- # 2️⃣ FILTER (kondisi logika) ## a) Filter satu kondisi
`df[df["gender_ketua_kelompok"] == "L"]`
## b) Multiple kondisi (`&`, `|`)
`df[ (df["provinsi"] == "JAWA BARAT") & (df["gender_ketua_kelompok"] == "P")]`
⚠️ WAJIB pakai tanda kurung! --- ## c) Filter dengan `isin`
`df[df["provinsi"].isin(["JAWA BARAT", "JAWA TENGAH"])]`
## d) Filter string (`contains`, `startswith`)
`df[df["nama_kelompok"].str.contains("NELAYAN", case=False, na=False)]`
--- ## e) Filter angka
`df[df["skor_kelompok"] >= 80]`
--- # 3️⃣ GROUPING (ringkas data) ## a) Group by satu kolom
`df.groupby("provinsi").size()`
Sama dengan:
`df["provinsi"].value_counts()`
--- ## b) Group + agregasi
`df.groupby("provinsi")["nama_kelompok"].nunique()`
--- ## c) Group by multiple kolom
`df.groupby(["provinsi", "gender_ketua_kelompok"]).size()`
--- ## d) Banyak agregasi sekaligus
`df.groupby("provinsi").agg( total_kelompok=("nama_kelompok", "count"), unik_kelompok=("nama_kelompok", "nunique"), rata_skor=("skor_kelompok", "mean"),)`
--- # 4️⃣ SUBSET + GROUP + FILTER (REAL CASE) ### ❓ Jumlah kelompok per provinsi, hanya yang > 1000
`( df.groupby("provinsi") .size() .reset_index(name="jumlah") .query("jumlah > 1000") .sort_values("jumlah", ascending=False))`
--- ### ❓ Distribusi gender per provinsi
`( df.groupby(["provinsi", "gender_ketua_kelompok"]) .size() .unstack(fill_value=0))`
--- ### ❓ Kelompok perempuan di Jawa Barat
`df[ (df["provinsi"] == "JAWA BARAT") & (df["gender_ketua_kelompok"] == "P")][["nama_kelompok", "kab", "desa_kelompok"]]`
--- # 5️⃣ FILTER DENGAN `query()` (lebih readable)
`df.query("provinsi == 'JAWA BARAT' and gender_ketua_kelompok == 'P'")`
Dengan `in`:
`df.query("provinsi in ['JAWA BARAT','JAWA TENGAH']")`
--- # 6️⃣ GROUP + FILTER (HAVING-style) Contoh: **provinsi dengan ≥ 100 kelompok perempuan**
`g = df[df["gender_ketua_kelompok"] == "P"]( g.groupby("provinsi") .size() .loc[lambda s: s >= 100])`
--- # 7️⃣ Best practice (penting banget) ### ✔️ Jangan lupa normalisasi dulu
`df["provinsi"] = df["provinsi"].str.strip().str.upper()df["gender_ketua_kelompok"] = df["gender_ketua_kelompok"].str.strip().str.upper()`
### ✔️ Hindari chained indexing ❌
`df[df["provinsi"] == "JAWA BARAT"]["kab"]`
✔️
`df.loc[df["provinsi"] == "JAWA BARAT", "kab"]`
--- ## TL;DR (pola paling sering)
`df.loc[ df["kolom"].isin([...]), ["kolom1","kolom2"] ]df.groupby([...]).agg(...)df.query("kondisi")`