overwriting values in dataframe in python

Question

I'm new to python so this might seem a bit easy problem for you expertise.

I've 6 categories(0 to 5),each of which has 4 sub-categories namely: '3','4','5','6'.

for this,I've created a dataframe using:

df=pd.DataFrame(index=list(range(5)),columns=list([3,4,5,6])

Now,I'm getting some calculated values from my loop:

for i in range(5):
    for j in list([3,4,5,6]):
       somecalculation=a

Now,I'm trying to replace the values of df with these calculations like for second iteration (i.e. for i=0,j=4), I got somecalculation=b, for third somecalculation=c and further d. When loop again iterates over i=2,I get calculations as e,f,g,h and so on for further iterations. I'm trying to append these values to df as soon as I obtain them but I'm not getting the desired output as

  3 4 5 6 
0 a b c d
1 e f g h
2 i j k l
.........
.........
.........

because ultimately,I want to take average of the column values using their indices, but replacing values of dataframe is becoming troublesome.

Siraj S. · Accepted Answer · 2016-07-12 21:37:50Z

1

your suggestion of appending to DataFrame rows iteratively is not optimal. it will slowdown the code. instead you can append the output to a list and than reshape the list as you like and ultimately convert it to pd.DataFrame. that will be way faster than what you propose. for example:

import pandas as pd, numpy as np
list1=[] #initialize the list
list1.extend([i]) #where i is the output from your loop
df = pd.DataFrame(np.reshape(list1,(5,4)),columns=['a','b','c','d']) # finally you can reshape the list as desired (eg: 5 row X 4 cols) and write it to dataframe in a single step

if this helps please do not forget to accept it:)

answered Jul 12, 2016 at 21:37

Siraj S.

3,7714 gold badges37 silver badges50 bronze badges

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

Comments

jnie · Accepted Answer · 2016-07-11 13:11:05Z

0

Not sure if I understand your question correctly, but there are multiple ways to insert a value to a desired dataframe cell.

For example line,

df.xs(1)['e'] = x

inserts value x to a cell that has an index value of 1 and is in column 'e'.

If you wan't to calculate the mean of the certain column and insert that to spesific cell, you could do:

import numpy as np

df.xs(desired_index_value)['desired_column_name'] = np.mean(df.xs('col_x',axis=1))

answered Jul 11, 2016 at 13:11

jnie

134 bronze badges

1 Comment

Tanvi Malhotra Over a year ago

I don't want to mention every time that I need to replace this or that value of the dataframe.I want my calculations to automatically replace the values of the df,once they are calculated in one iteration

Collectives™ on Stack Overflow

overwriting values in dataframe in python

2 Answers 2

Comments

1 Comment

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

1 Comment

Your Answer

Sign up or log in

Post as a guest

Related