CODEWITHSUNDEEP
pythonpython-3.xpandas
Oli
Oli

Reputation: 1371

How to set in pandas the first column and row as index?

When I read in a CSV, I can say pd.read_csv('my.csv', index_col=3) and it sets the third column as index.

How can I do the same if I have a pandas dataframe in memory? And how can I say to use the first row also as an index? The first column and row are strings, rest of the matrix is integer.

Upvotes: 100

Views: 268955

Answers (4)

develop_opi
develop_opi

Reputation: 43

Maybe try df = pd.read_csv(header = 0)

Upvotes: 0

Y. Yazarel
Y. Yazarel

Reputation: 1513

You can try this regardless of the number of rows

df = pd.read_csv('data.csv', index_col=0)

Upvotes: 110

villasv
villasv

Reputation: 6861

Making the first (or n-th) column the index in increasing order of verboseness:

df.set_index(list(df)[0])
df.set_index(df.columns[0])
df.set_index(df.columns.tolist()[0])

Making the first (or n-th) row the index:

df.set_index(df.iloc[0].values)

You can use both if you want a multi-level index:

df.set_index([df.iloc[0], df.columns[0]])

Observe that using a column as index will automatically drop it as column. Using a row as index is just a copy operation and won't drop the row from the DataFrame.

Upvotes: 45

Annie Guo
Annie Guo

Reputation: 319

Maybe try set_index()?

df = df.set_index([2])

Upvotes: 27

Related Questions