Introduction

To get the total of a Pandas column, use df["column"].sum(). That is the basic answer, but in real data work you also need to think about missing values, string-like numeric columns, grouping, and whether you want the result as a single scalar or as part of a report table.

The important thing is to make sure the column really contains numeric data and that Pandas is summing it with the semantics you expect. Otherwise the code can be short and still misleading.

The Basic Case

For a numeric column, .sum() gives the total directly.

1import pandas as pd
2
3df = pd.DataFrame({
4    "item": ["A", "B", "C"],
5    "sales": [100, 150, 200],
6})
7
8total = df["sales"].sum()
9print(total)

This returns 450.

That is the right answer for ordinary one-column totals.

Missing Values Are Ignored by Default

Pandas skips NaN values during summation unless you tell it not to.

df = pd.DataFrame({"sales": [100, None, 200]})
print(df["sales"].sum())
print(df["sales"].sum(skipna=False))

The first call returns 300.0. The second returns NaN.

That distinction matters when missing data means "unknown" rather than "zero or absent."

Make Sure the Column Is Actually Numeric

A common bug is that a column looks numeric but is really made of strings.

df = pd.DataFrame({"sales": ["100", "150", "200"]})
print(df["sales"].dtype)

Convert before summing:

df["sales"] = pd.to_numeric(df["sales"], errors="raise")
print(df["sales"].sum())

If you want invalid values to become missing rather than raising an error, use errors="coerce".

Conditional Totals

Often you want the total for only some rows.

1df = pd.DataFrame({
2    "region": ["East", "West", "East", "West"],
3    "sales": [100, 150, 200, 80],
4})
5
6east_total = df.loc[df["region"] == "East", "sales"].sum()
7print(east_total)

This is clearer than summing everything and then trying to subtract what you did not want.

Grouped Totals

If you need totals by category, use groupby().

grouped = df.groupby("region")["sales"].sum()
print(grouped)

You can group by more than one column too:

1df = pd.DataFrame({
2    "region": ["East", "East", "West", "West"],
3    "product": ["A", "B", "A", "B"],
4    "sales": [100, 200, 150, 80],
5})
6
7print(df.groupby(["region", "product"])["sales"].sum())

That returns one total per group combination rather than one overall scalar.

Summing Multiple Columns

If you want totals for several columns, select them first.

1df = pd.DataFrame({
2    "sales": [100, 150, 200],
3    "units": [5, 6, 7],
4})
5
6print(df[["sales", "units"]].sum())

This returns a Series with one total per selected column.

Add a Total Row for Reporting

Sometimes the total needs to live inside the final table rather than as a separate variable.

1df = pd.DataFrame({
2    "item": ["A", "B", "C"],
3    "sales": [100, 150, 200],
4})
5
6total_row = pd.DataFrame([{"item": "Total", "sales": df["sales"].sum()}])
7result = pd.concat([df, total_row], ignore_index=True)
8print(result)

This is useful for CSV exports, notebook displays, or quick internal reports.

Common Pitfalls

The biggest mistake is summing a column that only looks numeric but is actually stored as strings. Another is forgetting that .sum() skips missing values by default, which may not match the intended meaning of missing data. Developers also sometimes call .sum() on an entire mixed-type DataFrame without checking which columns are actually numeric. Finally, report-oriented code often needs grouped or filtered totals rather than one global scalar, so the shortest possible expression is not always the correct one.

Summary

• Use df["column"].sum() for the standard column total.
• Pandas skips NaN by default unless you set skipna=False.
• Convert string-like numeric columns before summing.
• Use filtering or groupby() when you need partial or grouped totals.
• Add a total row explicitly if the result needs to remain inside a report table.