pandas and NumPy 3 cheatsheet
pandas and NumPy 3: Grouping and combining
7 common pandas and NumPy 3 tasks.
Group and aggregate
Compute several statistics per group with named results.
df.groupby("country").agg(users=("id", "count"), avg_age=("age", "mean"))
Create a pivot table
Summarize values across two dimensions.
df.pivot_table(index="country", columns="plan", values="revenue", aggfunc="sum")
Join two DataFrames
Join on a key, like SQL. how can be inner, left, right or outer.
pd.merge(orders, users, left_on="user_id", right_on="id", how="left")
Compute a moving average
Average over a sliding window of seven rows.
df["avg_7"] = df["sales"].rolling(7).mean()
Aggregate by time period
Sum values by month. The index must be a datetime.
df.set_index("date")["sales"].resample("MS").sum()
Save to a CSV file
Write the DataFrame to a file without the row index.
df.to_csv("out.csv", index=False)