Top 10 pandas methods every analyst uses

Top 10 methods

1. 01

   read_csv

   df = pd.read_csv("sales.csv")

2. 02

   info

   df.info()

3. 03

   Filter rows

   df[df["amount"] > 500]

4. 04

   loc

   df.loc[df["city"] == "Pune", ["name", "amount"]]

5. 05

   groupby

   df.groupby("city")["amount"].sum()

6. 06

   sort_values

   df.sort_values("amount", ascending=False)

7. 07

   merge

   pd.merge(orders, users, on="user_id", how="left")

8. 08

   fillna

   df["amount"].fillna(0)

9. 09

   drop_duplicates

   df.drop_duplicates(subset="email")

10. 10

    value_counts

    df["city"].value_counts()

pandas: how do I …

1. 01

   Count rows and columns

   df.shape

2. 02

   See column types

   df.dtypes

3. 03

   Count missing values per column

   df.isna().sum()

4. 04

   Rename a column

   df.rename(columns={"amt": "amount"})

5. 05

   Drop a column

   df.drop(columns="notes")

6. 06

   Convert text to a date

   pd.to_datetime(df["date"])

7. 07

   Pull the month from a date

   df["date"].dt.month

8. 08

   Filter with two conditions

   df[(df.city == "Pune") & (df.amount > 500)]

9. 09

   Match a list of values

   df[df["city"].isin(["Pune", "Indore"])]

10. 10

    Search text in a column

    df[df["name"].str.contains("kumar", case=False)]

11. 11

    Count unique customers

    df["customer_id"].nunique()

12. 12

    Sum and mean per group

    df.groupby("city")["amount"].agg(["sum", "mean"])

13. 13

    Excel-style pivot

    df.pivot_table(index="city", columns="month",
        values="amount", aggfunc="sum")

14. 14

    Top 3 rows per city

    (df.sort_values("amount", ascending=False)
       .groupby("city").head(3))

15. 15

    Rank inside each group

    df.groupby("dept")["salary"].rank(ascending=False)

16. 16

    Previous row's value

    df["sales"].shift(1)

17. 17

    Month-on-month % change

    df["sales"].pct_change()

18. 18

    Running total

    df["amount"].cumsum()

19. 19

    New column from a condition

    np.where(df["amount"] > 1000, "High", "Low")

20. 20

    Stack two tables

    pd.concat([df1, df2])

21. 21

    Save to Excel

    df.to_excel("out.xlsx", index=False)

Two imports cover all of this: import pandas as pd and import numpy as np.