CODEWITHSUNDEEP
pythondataframeloopsfor-loop
bmaster69
bmaster69

Reputation: 131

For Loop for dataframe, using more efficient methods

I want to convert the column 1 with [i-1], if column 2 is False, iterated row by row. How can i do this in efficient way without using the for loop, example of for loop can be seen below:

import pandas as pd 

data_lamination = pd.DataFrame()
data_lamination['loss'] = [0.1,0.2,0.3,0.4,0.5]
data_lamination['nominal_wall_thickness'] = 1
data_lamination['flaw_distance'] = [10,0.2,0.3,0.4,0.01]
data_lamination['t_c'] = [0.1,0.01,0.3,0.01,0.5]
data_lamination['FCA'] = 0.1
data_lamination['lamination_distance_condition'] = data_lamination['flaw_distance'] >= 2 * data_lamination['t_c'] 

data_lamination['lamination_length']=[0.1,0.2,0.3,0.4,0.5]
data_lamination['lamination_depth']=[0.1,0.2,0.3,0.4,0.5]
    

new_value_length = []
new_value_depth = []

def abc(data_lamination, *args, **kwargs):
    

    for i, row in data_lamination.iterrows():
        if data_lamination.loc[i,'lamination_distance_condition'] == False:
        
            A = data_lamination.loc[i,'lamination_length'] + data_lamination.loc[i - 1,'lamination_length']
            B = data_lamination.loc[i,'lamination_depth'] + data_lamination.loc[i - 1,'lamination_depth']
            
        else:
             A = data_lamination.loc[i,'lamination_length']
             B = data_lamination.loc[i,'lamination_depth']
    
        new_value_length.append(A)
        new_value_depth.append(B)                    
    
    return A,B



B = abc(data_lamination)

data_lamination['lamination_length_organised'] = new_value_length 
data_lamination['lamination_depth_organised'] = new_value_depth

Your help will be much appriciated.

Thank you in advance.

Upvotes: 0

Views: 40

Answers (1)

Stefano
Stefano

Reputation: 274

you can define something like this

df['sum'] = df[['lamination_length']].apply(lambda x: [0] + [x.loc[i] + x.loc[i - 1] for i in range(1, len(x))])
df.loc[df['lamination_distance_condition'], 'new_col'] = df[df['lamination_distance_condition']]['lamination_length'].tolist()
df.loc[~df['lamination_distance_condition'], 'new_col'] = df[~df['lamination_distance_condition']]['sum'].tolist()

df is your data_lamination. So you define a temporary column that is length[i - 1] + length[i] (first element set to 0). if the condition is true, take from the old column, if false, take from the new one.

Edit: Maybe the final tolist() can be removed. I did not check.

Upvotes: 1

Related Questions