CODEWITHSUNDEEP
pythonpandasnumpy
Reut
Reut

Reputation: 1592

calculate LOG10 for few columns only in dataframe

I want to convert my given observations to log10.

This is the head of my DF: enter image description here

I want to calculate the log10 only for the fields from HR90 and left. I have tried to choose only the correct columns but I get error all the time and I thin my code is too much complex for this.

This is my code:

for i[9:] in data.columns:
 np.log10(data)

AttributeError: 'str' object has no attribute 'log10'

It seems like it still grabs also columns that I don't want to run this process on them. I have also tried to import math and then:

import math
for i in data.columns:
    if(data[i].dtype == np.float64 or data[i].dtype == np.int64):
        data.applymap(math.log10)

but thehn I have gotten:

TypeError: ('must be real number, not str', 'occurred at index NAME')

My end goal is to convert part of my observations to LOG10

Upvotes: 3

Views: 1801

Answers (3)

Nati
Nati

Reputation: 58

Do you mean something like this?

df = pd.DataFrame(np.random.randint(0,100,size=(100, 4)), columns=list('ABCD'))
>>>>
A   B   C   D
0   19  44  16  46
1   25  35  35  51
2   11  67  3   27
3   42  63  81  64
4   91  70  2   77

df[['A', 'B', 'C']] = df[['A', 'B', 'C']].apply(np.log10)
>>>
    A           B           C           D
0   0.106787    0.215757    0.080670    46
1   0.145489    0.188666    0.188666    51
2   0.017615    0.261519    -0.321371   27
3   0.210385    0.255113    0.280689    64
4   0.292044    0.266019    -0.521390   77
5   0.263046    0.223262    0.204679    63

This also will give you the same:

df[df.columns[:3]] = df[df.columns[:3]].apply(np.log10)

Upvotes: 2

CDJB
CDJB

Reputation: 14546

You can use get_loc to apply the logarithm map if the column is left or equal to 'HR90'.

for i in data.columns:
    if data.columns.get_loc(i) <= data.columns.get_loc('HR90'):
        data[i] = data[i].map(math.log10)

Upvotes: 0

hgrey
hgrey

Reputation: 3093

How about simply

data[data.columns[:9]] = np.log10(data[data.columns[:9]])

This will take log10 for all columns with index 0 to index 8 inclusive.

Upvotes: 1

Related Questions