Python Pandas Dataframes cheatsheet

 import pandas as pd

import numpy as np

# ═══════════════════════════════════════════════════════════════════════════════

# PANDAS DATAFRAME CHEATSHEET

# ═══════════════════════════════════════════════════════════════════════════════

# ── CREATING ──────────────────────────────────────────────────────────────────

pd.DataFrame({"A": [1,2], "B": [3,4]}) # from dict → most common

pd.DataFrame([{"A":1,"B":3}, {"A":2,"B":4}]) # from list of dicts

pd.DataFrame(np.array([[1,2],[3,4]]), columns=["A","B"]) # from NumPy array

pd.DataFrame({"A":[1,2]}, index=["r1","r2"]) # custom row index

# ── ATTRIBUTES ────────────────────────────────────────────────────────────────

# df.shape → (rows, cols)

# df.ndim → 2

# df.size → total elements

# df.dtypes → dtype per column

# df.columns → column labels

# df.index → row labels

# ── VIEWING ───────────────────────────────────────────────────────────────────

# df.head(n) → first n rows

# df.tail(n) → last n rows

# df.info() → dtypes + non-null counts

# df.describe() → count, mean, std, min, max

# ── SELECTING ─────────────────────────────────────────────────────────────────

# df["col"] → single column (Series)

# df[["col1","col2"]] → multiple columns (DataFrame)

# df.iloc[0] → row by position

# df.iloc[0:2] → rows by position slice

# df.loc[0] → row by label

# df.loc[0:1] → rows by label (inclusive)

# df.iloc[0, 1] → cell by position

# df.loc[0, "col"] → cell by label

# ── FILTERING ─────────────────────────────────────────────────────────────────

# df[df["Age"] > 26] → single condition

# df[(df["Age"] > 26) & (df["Score"] >= 90)] → multiple conditions (& = AND, | = OR)

# ── ADD / REMOVE COLUMNS ──────────────────────────────────────────────────────

# df["new"] = [...] → add column

# df["pass"] = df["Score"] >= 90 → derived column

# df.drop(columns=["col"], inplace=True) → remove column

# ── SORTING ───────────────────────────────────────────────────────────────────

# df.sort_values("col") → ascending

# df.sort_values("col", ascending=False) → descending

# ── MISSING VALUES ────────────────────────────────────────────────────────────

# df.isnull() → True where NaN

# df.isnull().sum() → count of NaN per column

# df.dropna() → drop rows with any NaN

# df.dropna(axis=1) → drop columns with any NaN

# df.dropna(how="all") → drop rows where ALL values are NaN

# df.dropna(thresh=2) → keep rows with at least 2 non-NaN values

# df.fillna(0) → fill NaN with 0

# df.fillna("") → fill NaN with empty string

# df.fillna("N/A") → fill NaN with custom label

# df.fillna({"col": val}) → fill per column

# df.fillna(method="ffill") → forward fill from previous row

# df.fillna(method="bfill") → backward fill from next row

# df.fillna(df.mean()) → fill with column mean

# ── GROUPBY ───────────────────────────────────────────────────────────────────

# df.groupby("col")["val"].mean() → mean per group

# df.groupby("col")["val"].agg(["min","max"]) → multiple aggregations

# df.groupby("col").size() → row count per group

# df.groupby(["col1","col2"])["val"].mean() → group by multiple columns

# ── UNIQUE ────────────────────────────────────────────────────────────────────

# df["col"].unique() → array of unique values

# df["col"].nunique() → count of unique values

# df.nunique() → unique count per column

# df["col"].value_counts() → frequency of each unique value

# ── MERGING & JOINING ─────────────────────────────────────────────────────────

# pd.merge(left, right) → inner join on common column (default)

# pd.merge(left, right, on="ID", how="inner") → inner : only matching rows

# pd.merge(left, right, on="ID", how="left") → left : all from left + matches

# pd.merge(left, right, on="ID", how="right") → right : all from right + matches

# pd.merge(left, right, on="ID", how="outer") → outer : all rows from both

# pd.merge(left, right, on=["col1","col2"]) → merge on multiple columns

# left.join(right) → join on INDEX (default: left join)

# ── CONCATENATING ─────────────────────────────────────────────────────────────

# pd.concat([df1, df2], ignore_index=True) → stack rows (axis=0)

# pd.concat([df1, df2], axis=1) → stack columns (axis=1)

# ── OPERATIONS ────────────────────────────────────────────────────────────────

# df["col"].apply(func) → apply function to each value

# df[cols].apply(func, axis=0) → apply function to each column

# df["col"].map(func) → transform each value in Series

# df.rename(columns={"old":"new"}) → rename columns

# df["col"].astype(float) → change dtype

# df.drop_duplicates() → remove duplicate rows

# df.set_index("col") → make column the index

# df.reset_index() → move index back to column

# df["col"].str.upper() → string methods via .str

# ── PIVOT TABLE ───────────────────────────────────────────────────────────────

# df.pivot_table(values="Salary", index="Dept") → mean per group

# df.pivot_table(values, index, columns="Level") → 2D cross-table

# df.pivot_table(values, index, aggfunc="sum") → sum

# df.pivot_table(values, index, aggfunc=["min","max","mean"]) → multiple aggs

# df.pivot_table(values, index, columns, fill_value=0) → fill NaN

# df.pivot_table(values, index, columns, margins=True) → add totals row/col

# ── MULTIINDEX ────────────────────────────────────────────────────────────────

# pd.MultiIndex.from_arrays([l1, l2], names=["A","B"]) → from lists

# pd.MultiIndex.from_tuples([(a,b),...], names=["A","B"])→ from tuples

# pd.MultiIndex.from_product([l1, l2], names=["A","B"]) → all combinations

# df.loc["HR"] → select level 0

# df.loc[("HR","Alice")] → select both levels

# df.xs("Alice", level="Name") → cross-section by value

# df.unstack("col") → move row level → columns

# df.stack("col") → move column level → rows

# ── DATA I/O ──────────────────────────────────────────────────────────────────

# df.to_csv("file.csv", index=False) → save CSV

# pd.read_csv("file.csv") → load CSV

# pd.read_csv("file.csv", usecols=["A","B"]) → load specific columns

# pd.read_csv("file.csv", nrows=100) → load first N rows

# df.to_json("file.json", orient="records") → save JSON

# pd.read_json("file.json") → load JSON

# df.to_excel("file.xlsx", index=False) → save Excel (needs openpyxl)

# pd.read_excel("file.xlsx") → load Excel

# ── ZIP (built-in) ────────────────────────────────────────────────────────────

# zip(l1, l2) → pairs elements → iterator of tuples

# list(zip(l1, l2)) → [(a1,b1), (a2,b2), ...]

# dict(zip(keys, values)) → {k:v, ...}

# pd.DataFrame(zip(l1,l2,l3), columns=[...]) → DataFrame from zipped lists

# zip(*pairs) → unzip list of tuples

# zip_longest(l1, l2, fillvalue="-") → zip with fill for unequal lengths

Pandas Dataframes Cheatsheet