create multiple new dataframes based on an existing data frames column in python

Question

I have a pandas data frame, df, that has 4 columns and a lot of rows.

I want to create 5 different data frames based on the value of one of the columns of the data frame. The column I am referring to is called color.

color has 5 unique values: red, blue, green, yellow, orange.

What I want to do is each of the 5 new data frames should contain all rows which have on of the values in color. For instance df_blue should have all the rows and columns where in the other data frame the value from the color column is blue.

The code I have is the following:

# create 5 new data frames
df_red = []
df_blue= []
df_green= []
df_yellow= []
df_orange= []
for i in range(len(df)):
    if df['color'] == "blue"
       df_blue.append(df)

# i would do if-else statements to satisfy all 5 colors

I feel I am missing some logic...any suggestions or comments?

Thanks!

DYZ · Accepted Answer · 2018-05-22 21:22:50Z

3

You need to use groupby. The following code fragment creates a sample DataFrame and converts it into a dictionary where colors are keys and the matching dataframes are values:

df = pd.DataFrame({'color': ['red','blue','red','green','blue'],
                   'foo': [1,2,3,4,5]})
colors = {color: dfc for color,dfc in df.groupby('color')}
#{'blue':   color  foo
#         1  blue    2
#         4  blue    5, 
# 'green':    color  foo
#          3  green    4, 
# 'red':   color  foo
#        0   red    1
#        2   red    3}

edited May 22, 2018 at 21:22

answered May 22, 2018 at 21:02

DYZ

57.3k10 gold badges73 silver badges101 bronze badges

Sign up to request clarification or add additional context in comments.

6 Comments

Scott Boston Over a year ago

Dictionary comprehension nice. +1

cool_beans Over a year ago

I get this error AttributeError: 'dict' object has no attribute 'groupby', also how can I do using data frames? I tried data.groupby('color') and it outputs <pandas.core.groupby.DataFrameGroupBy object at 0x000000003193DF28>. I want to have 5 data frames for each unique color.

DYZ Over a year ago

You said df is a DataFrame. Why is it a dict?

cool_beans Over a year ago

oh, the first error that says dict is the output of putting this code colors = {color: dfc for color,dfc in df.groupby('color')}

cool_beans Over a year ago

the second error i mentioned was a result of this code data.groupby('color')

|

cool_beans · Accepted Answer · 2018-05-22 21:48:09Z

-1

I ended up doing this for each of the colors.

  blue_data = data[data.color =='blue']

answered May 22, 2018 at 21:48

cool_beans

1411 gold badge5 silver badges15 bronze badges

3 Comments

jpp Over a year ago

This is not a good idea. You should use a dictionary for a variable number of variables.

jpp Over a year ago

It's inefficient, manual, hard to track, assumes you know your colours beforehand, prone to error, polluting the namespace, unstructured.

cool_beans Over a year ago

hmm i see...well my dataframe consists of pandas series, that is why i did it that way

Collectives™ on Stack Overflow

create multiple new dataframes based on an existing data frames column in python

2 Answers 2

6 Comments

3 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

6 Comments

3 Comments

Your Answer

Sign up or log in

Post as a guest

Related