How to Merge Two Columns in pandas (+, str.cat, agg)

To merge two columns in pandas, add them with + and a separator string. A sales team recently wanted one mailing-label column from separate city and state fields, and that one line did it.

It breaks when a ZIP code is an integer or a name is missing: you get a TypeError or a silent NaN.

Quick answer: add the columns with + and put a separator string between them. Convert non-text columns with astype(str) first:

df["location"] = df["city"] + ", " + df["state"]

In this tutorial, I’ll show you how to merge two columns pandas style on one customer table: text, numbers, missing values, three or more columns, and filling gaps.

How to Merge Two Text Columns in pandas with +

The running sample is four US customers, tested with Python 3.12.5 with pandas 3.0.6. For new-column basics, see my guide on adding a column to a pandas DataFrame covers the basics.

import pandas as pd

df = pd.DataFrame({
    "first_name": ["Emily", "James", "Olivia", "Marcus"],
    "last_name": ["Carter", "Nguyen", None, "Reed"],
    "city": ["Austin", "Denver", "Boston", "Miami"],
    "state": ["TX", "CO", "MA", "FL"],
    "zip": [73301, 80202, 2134, 33101],
})

df["location"] = df["city"] + ", " + df["state"]
df["full_name"] = df["first_name"] + " " + df["last_name"]
print(df[["location", "full_name"]])

Output:

     location     full_name
0  Austin, TX  Emily Carter
1  Denver, CO  James Nguyen
2  Boston, MA           NaN
3   Miami, FL   Marcus Reed

Merge a Text Column with a Number Column

pandas won’t add text to the integer zip column, so convert with astype(str). More conversions are in changing a column data type in pandas.

import pandas as pd

df = pd.DataFrame({
    "first_name": ["Emily", "James", "Olivia", "Marcus"],
    "last_name": ["Carter", "Nguyen", None, "Reed"],
    "city": ["Austin", "Denver", "Boston", "Miami"],
    "state": ["TX", "CO", "MA", "FL"],
    "zip": [73301, 80202, 2134, 33101],
})

try:
    df["state_zip"] = df["state"] + " " + df["zip"]
except TypeError as e:
    print("TypeError:", str(e).splitlines()[0][:70])

df["state_zip"] = df["state"] + " " + df["zip"].astype(str).str.zfill(5)
print(df[["state", "zip", "state_zip"]])

Output:

TypeError: can only concatenate str (not "int") to str
  state    zip state_zip
0    TX  73301  TX 73301
1    CO  80202  CO 80202
2    MA   2134  MA 02134
3    FL  33101  FL 33101

How Do You Merge Columns with Missing Values?

Use Series.str.cat() with na_rep="" (missing becomes empty text), then str.strip() the leftover space:

import pandas as pd

df = pd.DataFrame({
    "first_name": ["Emily", "James", "Olivia", "Marcus"],
    "last_name": ["Carter", "Nguyen", None, "Reed"],
    "city": ["Austin", "Denver", "Boston", "Miami"],
    "state": ["TX", "CO", "MA", "FL"],
    "zip": [73301, 80202, 2134, 33101],
})

df["full_name"] = df["first_name"].str.cat(df["last_name"], sep=" ", na_rep="")
df["full_name"] = df["full_name"].str.strip()
print(df[["first_name", "last_name", "full_name"]])

Output:

  first_name last_name     full_name
0      Emily    Carter  Emily Carter
1      James    Nguyen  James Nguyen
2     Olivia       NaN        Olivia
3     Marcus      Reed   Marcus Reed

Plain Python strings work differently; see concatenating strings in Python.

Merge Three or More Columns into One

agg(", ".join, axis=1) joins any number of string columns per row:

import pandas as pd

df = pd.DataFrame({
    "first_name": ["Emily", "James", "Olivia", "Marcus"],
    "last_name": ["Carter", "Nguyen", None, "Reed"],
    "city": ["Austin", "Denver", "Boston", "Miami"],
    "state": ["TX", "CO", "MA", "FL"],
    "zip": [73301, 80202, 2134, 33101],
})

df["zip5"] = df["zip"].astype(str).str.zfill(5)
df["address"] = df[["city", "state", "zip5"]].agg(", ".join, axis=1)
print(df["address"].tolist())

Output:

['Austin, TX, 73301', 'Denver, CO, 80202', 'Boston, MA, 02134', 'Miami, FL, 33101']

To break a merged column apart again, see splitting a pandas column by a delimiter.

Fill One Column’s Gaps from Another with combine_first()

To keep the mobile number and fall back to the home number only when mobile is blank, use combine_first():

import pandas as pd

df = pd.DataFrame({
    "name": ["Emily", "James", "Olivia", "Marcus"],
    "mobile": ["512-555-0142", None, "617-555-0199", None],
    "home": [None, "303-555-0117", "617-555-0100", None],
})
df["phone"] = df["mobile"].combine_first(df["home"])
print(df)

Output:

     name        mobile          home         phone
0   Emily  512-555-0142           NaN  512-555-0142
1   James           NaN  303-555-0117  303-555-0117
2  Olivia  617-555-0199  617-555-0100  617-555-0199
3  Marcus           NaN           NaN           NaN

Don’t confuse this with combining two tables. Joining DataFrames on a key is pandas merge and join, and stacking them is concatenating two DataFrames.

Common Errors When Merging pandas Columns

• TypeError with +: one column is numeric. Add astype(str).
• NaN in the result: a source value is missing. Use str.cat(na_rep="") or fillna("").
• ZIP like 2134.0: the column is float. Use astype("Int64") first.

When you’re cleaning up pandas columns, these walk-throughs come in handy:

• Replace values in a pandas column
• Set column names in pandas
• Get pandas column names
• Drop rows based on a column value

Frequently Asked Questions