Pandasdrillunique and nunique
df["col"].unique() returns an array of the distinct values in order of appearance and df["col"].nunique() returns how many there are, excluding NaN by default. df.nunique() gives the count for every column, and groupby("k")["col"].nunique() counts distinct values per group.
Syntax
Series.unique() | Series.nunique(dropna=True) | df.nunique(axis=0, dropna=True)| Part | What it is |
|---|---|
dropna | nunique: False counts NaN as a value |
Examples
import pandas as pd
df = pd.DataFrame({"cust": ["a", "b", "a", "c", None],
"region": ["N", "N", "S", "S", "S"]})
df["cust"].unique()Result: array(['a', 'b', 'c', None], dtype=object)
unique keeps None/NaN as a value.
df["cust"].nunique()Result: 3
df.groupby("region")["cust"].nunique()Result: region N 2 S 2 Name: cust, dtype: int64
South has a and c; the missing customer is not counted.
Common mistakes
- unique() returns a numpy array, not a list or Series: wrap with list() or use drop_duplicates() for a Series.
- nunique and len(unique()) differ by one: unique includes NaN, nunique excludes it unless dropna=False.
- Counting distinct across two columns with nunique only counts each column separately; use df[["a", "b"]].drop_duplicates().shape[0] for distinct pairs.
Practice unique and nunique
These drills check your answer on live data, free, no account needed.
- Count rows per groupReturn the number of orders each rep has, as a Series indexed by rep in alphabetical order.
- Remove duplicate recordsThe `leads` table has repeated emails. Keep only the most recent row for each email, which is the last one in the table. Keep the original row labels.
Questions
How do I get unique values sorted?
sorted(df["col"].dropna().unique()) or df["col"].drop_duplicates().sort_values().
How do I count distinct values with their frequencies?
value_counts() lists each distinct value with how often it occurs.
How do I list unique values of all columns?
{c: df[c].unique() for c in df.columns} builds a dict; df.nunique() gives just the counts.
Related
- value_countsHow to count values in a pandas column with value_counts: counts sorted by frequency, percentages with normalize=True, including NaN, and two-column combos.
- drop_duplicatesHow to remove duplicates in pandas with drop_duplicates: whole rows or a subset of columns, keep first, last or none, and finding duplicates with duplicated.
- groupbyHow groupby works in pandas: group by one or more columns, apply sum, mean, count or size, get a flat table with as_index=False, and iterate over groups.
- describeWhat df.describe returns in pandas: count, mean, std, min, quartiles and max for numeric columns, including text columns, custom percentiles and transposing.