How Do I extract Strings from a csv.file and write them as a list of strings

I would like to extract some strings from certain columns in a csv-file if one condition in another column is met. Then I want to write the extracted strings in a list in a txt.file.

I am new to pandas, so there is probably an obvious solution for this, but my file generated with the code below turns up empty. If I print my variable "extracted rows" in line 12 I only get this: "Series([], dtype: object)" Any ideas?

import pandas as pd

def process_csv(file_name):
    # Read the CSV file
    df = pd.read_csv(file_name)

    # Assuming the columns are named as 'Column5', 'Column4' and 'Column3'
    # Convert 'Column5' to numeric
    df['Column5'] = pd.to_numeric(df['Column5'], errors='coerce')

    # Extract rows where 'Column5' is >= 18
    extracted_rows = df[df['Column5'] >= 18]

    # Create new strings by concatenating 'Column4' and 'Column3' (which need to be reverse order in generated string for my purpose 
    combined_strings = extracted_rows['Column4'] + " " + extracted_rows['Column3']
    
    print(combined_strings)

    # Write the combined strings to a txt file
    with open('file.txt', 'w') as f:
        for item in combined_strings:
            f.write('%s\n' % item)

process_csv('file.csv')

UPDATE: Taking up a suggestion I worked with apply and tried to find a solution for cases in which rows in column five contained two numbers and '-'. But now I only get those rows out that actually contained '-'. Drives me a little crazy:

import pandas as pd

def process_csv(file_name):
    # Read the CSV file
    df = pd.read_csv(file_name)

    # Check if strings in column 5 contain '-'
    # If so split at '-' and take the first part
    # Otherwise, keep the original string
    df.iloc[:, 4] = df.iloc[:, 4].apply(lambda x: x.split('-')[0] if len(str(x)) > 3 and '-' in str(x) else x)

    # Convert column 5 to numeric, set invalid parsing as NaN
    df.iloc[:, 4] = pd.to_numeric(df.iloc[:, 4], errors='coerce')

    # Replace NaNs (resulted from invalid parsing) with a negative number
    df.iloc[:, 4].fillna(-1, inplace=True)

    # Extract rows where column 5 is >= 18
    extracted_rows = df[df.iloc[:, 4] >= 18]

    # Create new strings by concatenating column 4 and column 3
    combined_strings = extracted_rows.iloc[:, 3] + " " + extracted_rows.iloc[:, 2]

   print(combined_strings)
   Write the combined strings to a txt file
   with open('file.txt', 'w') as f:
        for item in combined_strings:
            f.write("%s\n" % item)

process_csv('file.csv')

You could use apply. For more info and documentation: (https://pandas.pydata.org/docs/reference/api/pandas.DataFrame.apply.html)

import pandas as pd

df = pd.DataFrame({'Col1': ['a', 'b', 'c'], 'Col2': ['a', 'b', 'e'], 'Col3': ['e', 'f', 'g']})


def do_something(row):
# In this function, the first input parameter is the "row"
# of the DataFrame, you could have more input parameters,
# but this could be quite complicated.
    if row['Col1'] == row['Col2']:
        return row['Col1'] + ' ' + row['Col3']
        

df.apply(do_something, axis=1)

The following is the output:

>>> 
0     a e
1     b f
2    None
dtype: object

You could of course redirect the output into part of your DataFrame by doing this:

df.loc[:, 'output'] = df.apply(do_something, axis=1)

Hope this helps!