CODEWITHSUNDEEP
pythonpandasdataframenumpy
almo
almo

Reputation: 561

Apply apply() function to multiple columns in pandas?

I used to have a column "combined" in a dataframe like this:

       combined

 0   0.97,0.92,0.65,0.24,0.15,0.19   
 1   0.96,0.86,0.69,0.38,0.56,0.78   
 2   0.64,0.69,0.93,0.07,0.73,0.05
 3   0.29,0.56,0.35,0.76,0.19,0.07

So when I tried to apply a numpy function frequency_changer(), I just need to do like this:

def frequency_changer(y):
    x_val_30 = [] 
    for i in range(0, len(y), 4):
        x_val_30.append(i)


    x_val = [] 
    for i in range(len(y)):  
        x_val.append(i)
    y_val = y
    cs = CubicSpline(x_val, y_val)
    x2 = cs(x_val_30)
    return x2


df['combined']=df['combined'].apply(frequency_changer)

Now, my new dataframe has many columns and each column has a single value.

df_new

      0      1       2       3       4       5  
0   0.97    0.92    0.65    0.24    0.15    0.19    
1   0.96    0.86    0.69    0.38    0.56    0.78    
2   0.64    0.69    0.93    0.07    0.73    0.05
3   0.29    0.56    0.35    0.76    0.19    0.07

I applied the same function, it did not work anymore. In the new dataframe, the same function reduced number of total rows. (I want to have a function that works on the length of each rows). How should I solve it? Thank you!

Upvotes: 0

Views: 156

Answers (1)

Akshay Sehgal
Akshay Sehgal

Reputation: 19312

Try using the following. This should give you the same output that you were getting from running the above function on the combined column.

def frequency_changer(y):
    x_val_30 = [] 
    for i in range(0, len(y), 4):
        x_val_30.append(i)


    x_val = [] 
    for i in range(len(y)):  
        x_val.append(i)
    y_val = y
    cs = CubicSpline(x_val, y_val)
    x2 = cs(x_val_30)
    return x2


#Original df
df['combined'].apply(frequency_changer)

0    [0.97, 0.15]
1    [0.96, 0.56]
2    [0.64, 0.73]
3    [0.29, 0.19]
Name: combined, dtype: object

#For new df
new_out = new_df.T.apply(frequency_changer).T  #<------
print(new_out)

      0     1
0  0.97  0.15
1  0.96  0.56
2  0.64  0.73
3  0.29  0.19

If you want to combine the 2 columns into one list like your function was behaving for the previous df, then simply use new_out.values.tolist() and store it in a new dataframe or as a new column.

Upvotes: 1

Related Questions