Create multiple dataframe using for loop in python 2.7

Question

I have a list of locations

["HOME", "Office", "SHOPPING"]

and a pandas data frame "DF"

Start_Location  End_Location    Date
OFFICE          HOME            3-Apr-15
OFFICE          HOME            3-Apr-15
HOME            SHOPPING    3-Apr-15
HOME            SHOPPING    4-Apr-15
HOME            SHOPPING    4-Apr-15
SHOPPING    HOME            5-Apr-15
SHOPPING    HOME            5-Apr-15
HOME            SHOPPING    5-Apr-15

I want to create 3 different data frames for HOME, Office, SHOPPING using for loop, but I am not able to do it.

I am new to python

Please help.

Thanks lucy

What's your question? You don't know how to write a for loop? — Stop harming Monica
– Stop harming Monica, Commented Feb 13, 2016 at 1:47
I know how to write for loop. My question is how to create 3 different data frames using for loop like df1=DF[DF.Start_Location==locations[0]]. Hope this helps — Lucy
– Lucy, Commented Feb 13, 2016 at 1:49
make your for loop a variable on the locations so that you do locations[i], and store them in a list of dataframe dfs[i] = — Zeugma
– Zeugma, Commented Feb 13, 2016 at 3:04

Lucy · Accepted Answer · 2016-02-14 17:14:58Z

5

I got the answer which I was looking for

import pandas as pd
gbl = globals()
for i in locations:
gbl['df_'+i] = df[df.Start_Location==i]

This will create 3 data frames df_HOME, df_office and df_SHOPPING

Thanks,

edited Feb 14, 2016 at 17:14

answered Feb 14, 2016 at 16:43

Lucy

4016 silver badges16 bronze badges

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

Comments

HYRY · Accepted Answer · 2016-02-14 22:31:28Z

Use groupby() and then call it's get_group() method:

import pandas as pd
import io

text = b"""Start_Location  End_Location    Date
OFFICE          HOME            3-Apr-15
OFFICE          HOME            3-Apr-15
HOME            SHOPPING    3-Apr-15
HOME            SHOPPING    4-Apr-15
HOME            SHOPPING    4-Apr-15
SHOPPING    HOME            5-Apr-15
SHOPPING    HOME            5-Apr-15
HOME            SHOPPING    5-Apr-15"""

locations = ["HOME", "OFFICE", "SHOPPING"]

df = pd.read_csv(io.BytesIO(text), delim_whitespace=True)
g = df.groupby("Start_Location")
for name, df2 in g:
    globals()["df_" + name.lower()] = df2

but I think add global variables in a for loop isn't a good method, you can convert the groupby to a dict by:

d = dict(iter(g))

then you can use d["HOME"] to get the data.

thanks for the solution, but I want to create these dfs without using read_csv, because the main dataframe is already available, and also if the list of location is more, lets say 20 then giving there name in left side of equal to sign will be little bit tidy. Is there any other way to do this?
read_csv() is only for demo, you don't need to call it. I edited the answer to use globals() .

Luis Fernando Larota · Accepted Answer · 2020-07-06 20:33:55Z

0

You could probably have a dictionary and you want to convert it to some dataframes, based on the keys of your dictionary:

gbl = globals()
for keys, values in dictionary.items():
   gbl['df_min'+ str(keys)] = pd.DataFrame(values)

answered Jul 6, 2020 at 20:33

Luis Fernando Larota

11 silver badge

Collectives™ on Stack Overflow

Create multiple dataframe using for loop in python 2.7

3 Answers 3

Comments

2 Comments

Comments

Linked

Hot Network Questions

Collectives™ on Stack Overflow

3 Answers 3

Comments

2 Comments

Comments

Linked

Related