I have this empty pandas dataframe and a function value(x,y) which takes in 2 arguments, the row number and the column number of the point in the pandas dataframe. I was wondering if there is a simpler way to iterate through the entire empty dataframe using these arguments, using df.apply perhaps.
I know it is possible to go through each individual column and run df.apply on the separate columns, but is it possible to do it without running any loops or what so ever.
Essentially, I am looking for something like this which I can run on the entire dataframe
df_copy.apply(lambda x: myfunction(x.value, x.column))
However, x.column does not exist, so it there is another way to do it or am I doing something wrong
Thanks!
Yes, use name and index attributes of a series:
df = pd.DataFrame(1, index = np.arange(10,51,10), columns = np.arange(5))
Show input dataframe
0 1 2 3 4
10 1 1 1 1 1
20 1 1 1 1 1
30 1 1 1 1 1
40 1 1 1 1 1
50 1 1 1 1 1
Let's define custom fuction and use rows as columns to do a calcuation.
def f(x):
#add row values to column values
return x.name + x.index
df.apply(f)
Output:
0 1 2 3 4
10 10 11 12 13 14
20 20 21 22 23 24
30 30 31 32 33 34
40 40 41 42 43 44
50 50 51 52 53 54
Note: apply is passing each column of the dataframe(which is a pd.Series) into the function f. Each series has an attribute name which is the column heading, and index, which is dataframe row index. So, function, f returns a calculated pd.Series for each column of the dataframe and is put back together as a dataframe.
Answering question in comments, let's use strings:
df = pd.DataFrame(1, index=['Ted','Bill','Ralph','John','Tim'], columns=['A','B','C','D','E'])
def f(x):
#Concatenate row values with column values
return x.index + '_' + x.name
df.apply(f)
OR use lambda function
df.apply(lambda x: x.index + '_' + x.name)
Output:
A B C D E
Ted Ted_A Ted_B Ted_C Ted_D Ted_E
Bill Bill_A Bill_B Bill_C Bill_D Bill_E
Ralph Ralph_A Ralph_B Ralph_C Ralph_D Ralph_E
John John_A John_B John_C John_D John_E
Tim Tim_A Tim_B Tim_C Tim_D Tim_E