CODEWITHSUNDEEP
pythonpandas
Thegamer23
Thegamer23

Reputation: 567

Order dataframe by row

Suppose I have the following dataframe

      C1  C2
John  4   3
Bob   5   7
Mary  6   5
Carl  5   6
James 4   3

How can I order my dataframe to have:

Carl, Mary, Bob, John, James

in an efficient way?

It is an arbitrary ordering and I might have the names stored in a variable orderedNames

Upvotes: 1

Views: 283

Answers (2)

piRSquared
piRSquared

Reputation: 294546

define your ordering 

arbitrary_ordering = ['Carl', 'Mary', 'Bob', 'John', 'James']

option 1
loc 

df.loc[arbitrary_ordering]

option 2
reindex 

df.reindex(arbitrary_ordering)

option 3
reindex_axis 

df.reindex_axis(arbitrary_ordering)

All yield

       C1  C2
Carl    5   6
Mary    6   5
Bob     5   7
John    4   3
James   4   3

alternative 1 

df.iloc[df.index.to_series().map(arbitrary_ordering.index)]

alternative 2
pd.Categorical 

df.index = pd.Categorical(df.index, categories=arbitrary_ordering)
df.sort_index()

Time Test

enter image description here

Upvotes: 2

matusko
matusko

Reputation: 3777

You can order them arbirtrarly using loc (suppose your data are in df)

df.loc[['Carl', 'Mary', 'Bob', 'John', 'James'], :]

However, if you want to order them by subset of columns use sort_values

df.sort_values(by='C1')

Or even more advanced would be sorting by custom key function. See this question: pandas sort lambda function

Upvotes: 2

Related Questions