Pandas couple of questions

I have a task to do where I need to process 6.4k rows x 75 cols Excel sheet. I have to arrange the columns according to the recommendations, for eg.: Columns DE from input file will be merged into Column A in output file/spreadsheet. There will be 83 operations on whole columns performed. Some of the tasks to be performed:

  1. merging two columns into single one,
  2. replacing values in a column based on data in input file comparing them for eg.: Input File: Baltimore, Output file Games Shop BC
  3. Some computing, for eg.: Creating CustomAccountNumbers based on ClientID + some numbers

Hence my questions:

  • Will it be better to do it step by step for eg.: >Load Columns DE into dataframe from input file merge them into into single column and save as Column A in output file/spreadsheet< OR >Load whole excel file into dataframe1, create empy dataframe2, perform operations and step by step filling dataframe2 and then save dataframe2 as new file/spreadsheet<
  • I find fast solution for merging columns but it's not working in current pandas version. Could someone write some explanation for it and help me updating it for current pandas verion please? It's not working because of df.ix() has been deprecated. I tried replacing it with df.iloc() but I couldnt make it work. Or maybe there's something better for merging task?

    df[0][:, 1:].T.values, sep=' ')

For now I've created interface to read specific columns from Excel file based on user input and save them in correct order (I've chosen the first method (dealing with file step by step) but I don't know if it's the right way to do it) and I have stopped at the first task where I started to wonder if am I doing it correctly and I decided to ask for Your help. I know that I've made very long text but I don't have anyone to ask these questions. I will be grateful for everyone who replies. Alse if someone could point me at some tools for task 2 and 3 it would be great.

Best regards, Ed.