CODEWITHSUNDEEP
pythonsearchpandasdataframe
Amanda
Amanda

Reputation: 885

replace column values in one dataframe by values of another dataframe

I have two dataframes, the first one has 1000 rows and looks like:

Date            Group         Family       Bonus
2011-06-09      tri23_1       Laavin       456
2011-07-09      hsgç_T2       Grendy       679
2011-09-10      bbbj-1Y_jn    Fantol       431
2011-11-02      hsgç_T2       Gondow       569

The column Group has different values, sometimes repeated, but in general about 50 unique values.

The second dataframe contains all these 50 unique values (50 rows) and also the hotels, that are associated to these values:

Group             Hotel
tri23_1           Jamel
hsgç_T2           Frank
bbbj-1Y_jn        Luxy
mlkl_781          Grand Hotel
vchs_94           Vancouver

My goal is to replace the value in the column Group of the first dataframe by the corresponding values of the column Hotel of the second dataframe/or create the column Hotel with the corresponding values. When I try to make it just by assignment like

df1.loc[(df1.Group=df2.Group), 'Hotel']=df2.Hotel

I have an error that the dataframes are not of equal size, so the comparison is not possible.

Upvotes: 49

Views: 114138

Answers (5)

Greg Friedman
Greg Friedman

Reputation: 341

You could also create a dictionary and use map:

hotel_dict = df2.set_index('Group').to_dict()
df1['Group'] = df1['Group'].map(lambda x: hotel_dict[x])

Upvotes: 11

vkt
vkt

Reputation: 509

This is an old question but here is another way to do it, it is not like the pandas way but is fast

Reproducing the dataframe 1 - this is to be updated

df_1

    Date    Group   Family  Bonus
0   2011-06-09  tri23_1     Laavin  456
1   2011-07-09  hsgç_T2     Grendy  679
2   2011-09-10  bbbj-1Y_jn  Fantol  431
3   2011-11-02  hsgç_T2     Gondow  569

Reproducing dataframe 2 - the look up

df_2

    Group   Hotel
0   tri23_1     Jamel
1   hsgç_T2     Frank
2   bbbj-1Y_jn  Luxy
3   mlkl_781    Grand Hotel
4   vchs_94     Vancouver

Get all the hotel id (key column) from the dataframe 1 as a list

key_list = list(df_1['Group'])

['tri23_1', 'hsgç_T2', 'bbbj-1Y_jn', 'hsgç_T2']

Create a dictionary from the look up dataframe which has the key col and the value col

dict_lookup = dict(zip(df_2['Group'], df_2['Hotel']))

{'bbbj-1Y_jn': 'Luxy',
 'hsgç_T2': 'Frank',
 'mlkl_781': 'Grand Hotel',
 'tri23_1': 'Jamel',
 'vchs_94': 'Vancouver'}

Replace the value by creating a list by looking up the value and assign to dataframe 1 column

df_1['Group'] = [dict_lookup[item] for item in key_list]

Updated dataframe 1

    Date    Group   Family  Bonus
0   2011-06-09  Jamel   Laavin  456
1   2011-07-09  Frank   Grendy  679
2   2011-09-10  Luxy    Fantol  431
3   2011-11-02  Frank   Gondow  569

Upvotes: 7

Erik Christiansen
Erik Christiansen

Reputation: 1

Columns in pandas DataFrames are just Series. Make the DataFrames (or DataFrame and Series, as shown here) share the same index so that assignment can occur from the Series to the DataFrame:

**In:**

df = pd.DataFrame(data=
{'date': ['2011-06-09', '2011-07-09', '2011-09-10', '2011-11-02'], 
'family': ['Laavin', 'Grendy', 'Fantol', 'Gondow'], 
'bonus': ['456', '679', '431', '569']}, 
index=pd.Index(name='Group', data=['tri23_1', 'hsgç_T2', 'bbbj-1Y_jn', 'hsgç_T2']))

**Out:**
            date    family  bonus
Group           
tri23_1 2011-06-09  Laavin  456
hsgç_T2 2011-07-09  Grendy  679
bbbj-1Y_jn  2011-09-10  Fantol  431
hsgç_T2 2011-11-02  Gondow  569

**In:**

hotel_groups = pd.Series(['Jamel', 'Frank', 'Luxy', 'Grand Hotel', 'Vancouver'], 
index=pd.Index(name='Group', data=['tri23_1', 'hsgç_T2', 'bbbj-1Y_jn', 'mlkl_781', 'vchs_94']))

**Out:**

Group
tri23_1             Jamel
hsgç_T2             Frank
bbbj-1Y_jn           Luxy
mlkl_781      Grand Hotel
vchs_94         Vancouver
dtype: object

**In:**

df['hotel'] = hotel_groups

**Out:**

            date    family  bonus hotel
Group               
tri23_1 2011-06-09  Laavin  456 Jamel
hsgç_T2 2011-07-09  Grendy  679 Frank
bbbj-1Y_jn  2011-09-10  Fantol  431 Luxy
hsgç_T2 2011-11-02  Gondow  569 Frank

Notice that the index of both is 'Group', which allows the assignment.

If you assign a like-indexed Series to a DataFrame column, the assignment works. Notice that this works despite there being duplicate group values in df. It would not work if there were duplicate index values (with different corresponding data values) in the hotel_groups Series (e.g., if there were two entries for index value hsgc_T2, the first with data value Frank and the second with data value Luxy that is being assigned to df['hotel'] (not that this would ever occur in your example). This wouldn't work because there wouldn't be a way to know which value to assign the like-indexed DataFrame column.

Upvotes: 0

176coding
176coding

Reputation: 3059

just use pandas join, you can refer to detail link: http://pandas.pydata.org/pandas-docs/stable/merging.html

df1.join(df2,on='Group')

Upvotes: 4

EdChum
EdChum

Reputation: 394031

If you set the index to the 'Group' column on the other df then you can replace using map on your original df 'Group' column:

In [36]:
df['Group'] = df['Group'].map(df1.set_index('Group')['Hotel'])
df

Out[36]:
         Date  Group  Family  Bonus
0  2011-06-09  Jamel  Laavin    456
1  2011-07-09  Frank  Grendy    679
2  2011-09-10   Luxy  Fantol    431
3  2011-11-02  Frank  Gondow    569

Upvotes: 76

Related Questions