Storing individual dataframes in a list using a loop Python

Question

I need to store 10 data frames into a list. The data frames are the outputs of a function. The problem is that the list remains all the same as the last data frame. For example, resultlist[0] suppose to be the data frame of 01-Feb-2020. But every data frame in the list is the output from 10-Feb-2020.

resultList = [None]*10

for i in range(0,10):    
    resultList[i]= getResultfunction(id,listDates[i])

emilioho2020 · Accepted Answer · 2020-04-17 12:11:45Z

1

As far as I can see, the code snippet you provided is correct. Just know that you are giving the same id parameter to each getResultfunction call. Could you show the implementation of the getResultfunction? The problem is probably in there. Either that or your listDates list contains the wrong dates.

EDIT 1 Alright so the problem is probably that you are just storing references to the same dataframe in the list. If you are not familiar with how programming languages store references and values, I'd suggest that you check out this link https://blog.penjee.com/passing-by-value-vs-by-reference-java-graphical/ or just google it, as its a fundamental paradigm of programming languages.

So basically, each time you call getResultfunction, you get a reference to the same template, then you alter this template, and return a reference to the template. Next time you call the function, you access the same (altered) template, change it again, and return a new reference to it. The list then stores lots of references to the same object!

In this particular case, each call to getResultfunction should create a new object. I see 2 solutions; Each time you call getResultfunction you:

just copy the existing template, which generates a new identical object:

def getResultfunction(id, date):
    p = template.copy()
    """other computations"""
    return p

create a new template from scratch:

def getResultfunction(id, date):
    data = {...}
    df = pd.Dataframe(data)
    """"other computations"""
    return df

edited Apr 17, 2020 at 12:11

answered Apr 16, 2020 at 12:27

emilioho2020

1029 bronze badges

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

5 Comments

user13321166 Over a year ago

HI, Emilio thank you for you answer. I just checked again. So I did it manually.. resultList[0]= getResultfunction(id,listDates[0]) resultList[1]= getResultfunction(id,listDates[1]) ... The last dataframe will overwrite every item in the list. Is there another way to store individual dataframe in a different data structure which will work?

emilioho2020 Over a year ago

It really depends on what type you are storing in the list. You are probably storing references to the same dataframe in the list and then just changing the same dataframe over and over again. Can you provide the code of the getResultfunction?

user13321166 Over a year ago

i am storing a panda df into the list. the function is similar to this: def getResultFucntion(id,date): p = template p.loc[0,1] = date p.loc[0,5] = id ... return p

emilioho2020 Over a year ago

I edited my post, I think this should solve your problem :)

user13321166 Over a year ago

hey emilio thank you for your help! The problem is solved!! : )

Collectives™ on Stack Overflow

Storing individual dataframes in a list using a loop Python

1 Answer 1

5 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

5 Comments

Your Answer

Sign up or log in

Post as a guest

Related