Develop Reference

How do you create new column from two distinct categorical column values in a dataframe by same column ID in pandas?

Sorry for the confusing title. I am practicing how to manipulate dataframes in Python through pandas. How do I make this kind of table:
id role name
0 11 ACTOR Luna Wedler, Jannis Niewöhner, Milan Peschel, ...
1 11 DIRECTOR Christian Schwochow
2 22 ACTOR Guy Pearce, Matilda Anna Ingrid Lutz, Travis F...
3 22 DIRECTOR Andrew Baird
4 33 ACTOR Glenn Fredly, Marcello Tahitoe, Andien Aisyah,...
5 33 DIRECTOR Saron Sakina
Into this kind:
id director actors name
0 11 Christian Schwochow Luna Wedler, Jannis Niewöhner, Milan Peschel, ...
1 22 Andrew Baird Guy Pearce, Matilda Anna Ingrid Lutz, Travis F...d
2 33 Saron Sakina Glenn Fredly, Marcello Tahitoe, Andien Aisyah,...

Try this way
df.pivot(index='id', columns='role', values='name')

You can do in addition to #Tejas's answer:
df = (df.pivot(index='id', columns='role', values='name').
reset_index().
rename_axis('',axis=1).
rename(columns={'ACTOR':'actors name','DIRECTOR':'director'}))

Remove unwanted parts from strings in Dataframe

I am looking for an efficient way to remove unwanted parts from strings in a DataFrame column.
My dataframe:
Passengers
1 Sally Muller, President, Mark Smith, Vicepresident, John Doe, Chief of Staff
2 Sally Muller, President, Mark Smith, Vicepresident
3 Sally Muller, President, Mark Smith, Vicepresident, John Doe, Chief of Staff
4 Mark Smith, Vicepresident, John Doe, Chief of Staff, Peter Parker, Special Effects
5 Sally Muller, President, John Doe, Chief of Staff, Peter Parker, Special Effects, Lydia Johnson, Vice Chief of Staff
...
desired form of df:
Passengers
1 Sally Muller, Mark Smith, John Doe
2 Sally Muller, Mark Smith
3 Sally Muller, Mark Smith, John Doe
4 Mark Smith, John Doe, Peter Parker
5 Sally Muller, John Doe, Peter Parker, Lydia Johnson
...
Up to now I did it with endless handmade copy/paste regex list:
df = df.replace(r'President,','', regex=True)
df = df.replace(r'Vicepresident,','', regex=True)
df = df.replace(r'Chief of Staff,','', regex=True)
df = df.replace(r'Special Effects,','', regex=True)
df = df.replace(r'Vice Chief of Staff,','', regex=True)
...
Is there a more comfortable way to do this?
Edit
More accurate example of original df:
Passengers
1 Sally Muller, President, EL Mark Smith, John Doe, Chief of Staff, Peter Gordon, Director of Central Command
2 Sally Muller, President, EL Mark Smith, Vicepresident
3 Sally Muller, President, EL Mark Smith, Vicepresident, John Doe, Chief of Staff, Peter Gordon, Dir CC
4 Mark Smith, Vicepresident, John Doe, Chief of Staff, Peter Parker, Special Effects
5 President Sally Muller, John Doe Chief of Staff, Peter Parker, Special Effects, Lydia Johnson , Vice Chief of Staff
...
desired form of df:
Passengers
1 Sally Muller, Mark Smith, John Doe, Peter Gordon
2 Sally Muller, Mark Smith
3 Sally Muller, Mark Smith, John Doe, Peter Gordon
4 Mark Smith, John Doe, Peter Parker
5 Sally Muller, John Doe, Peter Parker, Lydia Johnson
...
Up to now I did it with endless handmade copy/paste regex list:
df = df.replace(r'President','', regex=True)
df = df.replace(r'Director of Central Command,','', regex=True)
df = df.replace(r'Dir CC','', regex=True)
df = df.replace(r'Vicepresident','', regex=True)
df = df.replace(r'Chief of Staff','', regex=True)
df = df.replace(r'Special Effects','', regex=True)
df = df.replace(r'Vice Chief of Staff','', regex=True)
...
messy output is like:
Passengers
1 Sally Muller, , Mark Smith, John Doe, , Peter Gordon,
2 Sally Muller, Mark Smith,
3 Sally Muller, Mark Smith,, John Doe, Peter Gordon
4 Mark Smith, John Doe, Peter Parker
5 Sally Muller,, John Doe, Peter Parker , Lydia Johnson,
...

If every passenger has their title, then you can use str.split + explode, then select every second item starting from the first item, then groupby the index and join back:
out = df['Passengers'].str.split(',').explode()[::2].groupby(level=0).agg(', '.join)
or str.split + explode and apply a lambda that does the selection + join
out = df['Passengers'].str.split(',').apply(lambda x: ', '.join(x[::2]))
Output:
0 Sally Muller, Mark Smith, John Doe
1 Sally Muller, Mark Smith
2 Sally Muller, Mark Smith, John Doe
3 Mark Smith, John Doe, Peter Parker
4 Sally Muller, John Doe, Peter Parker, Lydia...
Edit:
If not everyone has a title, then you can create a set of titles and split and filter out the titles. If the order of the names don't matter in each row, then you can use set difference and cast each set to a list in a list comprehension:
titles = {'President', 'Vicepresident', 'Chief of Staff', 'Special Effects', 'Vice Chief of Staff'}
out = pd.Series([list(set(x.split(', ')) - titles) for x in df['Passengers']])
If order matters, then you can use a nested list comprehension:
out = pd.Series([[i for i in x.split(', ') if i not in titles] for x in df['Passengers']])

This is one case where apply is actually faster that explode:
df2 = df['Passengers'].apply(lambda x: ', '.join(x.split(', ')[::2])) #.to_frame() # if dataframe needed
output:
Passengers
0 Sally Muller, Mark Smith, John Doe
1 Sally Muller, Mark Smith
2 Sally Muller, Mark Smith, John Doe
3 Mark Smith, John Doe, Peter Parker
4 Sally Muller, John Doe, Peter Parker, Lydia Jo...

We can create a full regex pattern match on every string you need to remove and replace.
This can handle situations were the passengers will not have a title.
df2 = df['Passengers'].str.replace("(President)|(Vicepresident)|(Chief of Staff)|(Special Effects)|(Vice Chief of Staff)", "",regex=True).replace("( ,)", "", regex=True).str.strip().str.rstrip(",")

Analyzing Token Data from a Pandas Dataframe

I'm a relative python noob and also new to natural language processing (NLP).
I have dataframe containing names and sales. I want to: 1) break out all the tokens and 2) aggregate sales by each token.
Here's an example of the dataframe:
name sales
Mike Smith 5
Mike Jones 3
Mary Jane 4
Here's the desired output:
token sales
mike 8
mary 4
Smith 5
Jones 3
Jane 4
Thoughts on what to do? I'm using Python.

Assumption: you have a function tokenize that takes in a string as input and returns a list of tokens
I'll use this function as a tokenizer for now:
def tokenize(word):
return word.casefold().split()
Solution
df.assign(tokens=df['name'].apply(tokenize)).explode('tokens').groupby('tokens')['sales'].sum().reset_index()
In [45]: df
Out[45]:
name sales
0 Mike Smith 5
1 Mike Jones 3
2 Mary Jane 4
3 Mary Anne Jane 1
In [46]: df.assign(tokens=df['name'].apply(tokenize)).explode('tokens').groupby('tokens')['sales'].sum().reset_index()
Out[46]:
tokens sales
0 anne 1
1 jane 5
2 jones 3
3 mary 5
4 mike 8
5 smith 5
Explanation
The assign step creates a column called tokens that applies the tokenize functio
Note: For this particular tokenize function - you can use df['name'].str.lower().str.split() - however this won't generalize to custom tokenizers hence the .apply(tokenize)
this generates a df that looks like
name sales tokens
0 Mike Smith 5 [mike, smith]
1 Mike Jones 3 [mike, jones]
2 Mary Jane 4 [mary, jane]
3 Mary Anne Jane 1 [mary, anne, jane]
use df.explode on this to get
name sales tokens
0 Mike Smith 5 mike
0 Mike Smith 5 smith
1 Mike Jones 3 mike
1 Mike Jones 3 jones
2 Mary Jane 4 mary
2 Mary Jane 4 jane
3 Mary Anne Jane 1 mary
3 Mary Anne Jane 1 anne
3 Mary Anne Jane 1 jane
last step is just a groupy-agg step.

You can use the str.split() method and keep item 0 for the first name, using that as the groupby key and take the sum, then do the same for item -1 (last name) and concatenate the two.
import pandas as pd
df = pd.DataFrame({'name': {0: 'Mike Smith', 1: 'Mike Jones', 2: 'Mary Jane'},
'sales': {0: 5, 1: 3, 2: 4}})
df = pd.concat([df.groupby(df.name.str.split().str[0]).sum(),
df.groupby(df.name.str.split().str[-1]).sum()]).reset_index()
df.rename(columns={'name':'token'}, inplace=True)

df[["fname", "lname"]] = df["name"].str.split(expand=True) # getting tokens,considering separated by space
tokens_df = pd.concat([df[['fname', 'sales']].rename(columns = {'fname': 'tokens'}),
df[['lname', 'sales']].rename(columns = {'lname': 'tokens'})])
pd.DataFrame(tokens_df.groupby('tokens')['sales'].apply(sum), columns=['sales'])

Find and replace in dataframe from another dataframe

I have two dataframes, here are snippets of both below. I am trying to find and replace the artists names in the second dataframe with the id's in the first dataframe. Is there a good way to do this?
id fullName
0 1 Colin McCahon
1 2 Robert Henry Dickerson
2 3 Arthur Dagley
Artists
0 Arthur Dagley, Colin McCahon, Maria Cruz
1 Fiona Gilmore, Peter Madden, Nicholas Spratt, ...
2 Robert Henry Dickerson
3 Steve Carr
Desired output:
Artists
0 3, 1, Maria Cruz
1 Fiona Gilmore, Peter Madden, Nicholas Spratt, ...
2 2
3 Steve Carr

You mean check with replace
df1.Artists.replace(dict(zip(df.fullName,df.id.astype(str))),regex=True)
0 3, 1, Maria Cruz
1 Fiona Gilmore, Peter Madden, Nicholas Spratt, ...
2 2
3 Steve Carr
Name: Artists, dtype: object

Convert your first dataframe into a dictionary:
d = Series(name_df.id.astype(str),index=name_df.fullName).to_dict()
Then use .replace():
artists_df["Artists"] = artists_df["Artists"].replace(d, regex=True)

Concatenate a set of column values based on another column in Pandas

Given a Pandas dataframe which has a few labeled series in it, say Name and Villain.
Say the dataframe has values such:
Name: {'Batman', 'Batman', 'Spiderman', 'Spiderman', 'Spiderman', 'Spiderman'}
Villain: {'Joker', 'Bane', 'Green Goblin', 'Electro', 'Venom', 'Dr Octopus'}
In total the above dataframe has 2 series(or columns) each with six datapoints.
Now, based on the Name, I want to concatenate 3 more columns: FirstName, LastName, LoveInterest to each datapoint.
The result of which adds 'Bruce; Wayne; Catwoman' to every row which has Name as Batman. And 'Peter; Parker; MaryJane' to every row which has Name as Spiderman.
The final result should be a dataframe containing 5 columns(series) and 6 rows each.

This is a classic inner-join scenario. In pandas, use the merge module-level function:
In [13]: df1
Out[13]:
Name Villain
0 Batman Joker
1 Batman Bane
2 Spiderman Green Goblin
3 Spiderman Electro
4 Spiderman Venom
5 Spiderman Dr. Octopus
In [14]: df2
Out[14]:
FirstName LastName LoveInterest Name
0 Bruce Wayne Catwoman Batman
1 Peter Parker MaryJane Spiderman
In [15]: pd.DataFrame.merge(df1,df2,on='Name')
Out[15]:
Name Villain FirstName LastName LoveInterest
0 Batman Joker Bruce Wayne Catwoman
1 Batman Bane Bruce Wayne Catwoman
2 Spiderman Green Goblin Peter Parker MaryJane
3 Spiderman Electro Peter Parker MaryJane
4 Spiderman Venom Peter Parker MaryJane
5 Spiderman Dr. Octopus Peter Parker MaryJane