Python- Automating Variable naming for with imported CSV data

Question

I'm a little new to python, stuck in on the 6.00x (spring 2013) course. I'd hoped to try some of my new found knowledge but appear to have overreached.

The idea was to import a load a CSV file containing my bank statement into python. I'd then hoped place turn each transaction into an instance of a class. I'd then hoped to start playing around with the data to see what I could do but I appear to be failing at even the first hurdle, getting things nicely fitted into my Object Oriented program.

I started with this to import my file:

import csv
datafile = open('PATH/TO/file.csv', 'r')
datareader = csv.reader(datafile)
data = []
for row in datareader:
    data.append(row)

That seems to work. I get a list of all the statement data that looks something like this below (You'll understand me not uploading the real data...)

[['date', 'type', 'details', 'amount', 'balance', 'accountname', 'accountdetails', 'blank_string'],['date', 'type', 'details', 'amount', 'balance', 'accountname', 'accountdetails', 'blank_string'],['date', 'type', 'details', 'amount', 'balance', 'accountname', 'accountdetails', 'blank_string'],['date', 'type', 'details', 'amount', 'balance', 'accountname', 'accountdetails', 'blank_string']]

so typing data[0] would get me:

['date', 'type', 'details', 'amount', 'balance', 'accountname', 'accountdetails', 'blank_string']

So then I created my class and constructor. With the idea of decomposing each one of these transactions into an easily accessible item.

class Transaction(object):
    """
    Abstract class for building different kinds of transaction
    """
    def __init__(self, data):
        self.date = data[0]
        self.trans_type = data[1]
        self.description = data[2]
        self.amount = data[3]
        self.balance = data[4]
        self.account_type = data[5]
        self.account_details = data[6]

I find this works if I now enter

T1 = Transaction(data[0])

However I don't want to have to constantly type T1 =... T2=... t3=... t4=... there's LOADS of transctions it'd take forever!

So I tried this!

for i in range(len(data)):
eval("T" + str(i)) = Transaction(data[i])

But python really doesn't like that... It reports back:

SyntaxError: There is an error in your program: * can't assign to function call (FILENAME.py, line 80)

So my question is. Why can't I iteratively use the eval() function to assign my data as an instance into class Transaction(object)?

If there's no way around that, how might else I go about doing so?

I also have a lingering doubt that my mistake his suggests I've missed some point about Object Orientated Programming and when its appropriate to use it. Would I be better just feeding my csv data into a dictionary as its imported and playing around with it from there?

Many thanks! Huw

Martijn Pieters · Accepted Answer · 2013-04-02T14:57:48.393

4

Use a transactions = [] list instead, simply .append() new Transaction instances:

transactions = []
for row in datareader:
    transactions.append(Transaction(row))

or even:

transactions = [Transaction(row) for row in datareader]

There is no need to create individual variables for each row result.

edited Apr 02 '13 at 14:57

answered Apr 02 '13 at 14:44

Martijn Pieters

1,048,767
296
4,058
3,343

1

And this appending could be done with list comprehension: `transactions = [Transaction(item) for item in data]` with no need of any `for i in range(len(data))` stuff ;) – Piotr Hajduga Apr 02 '13 at 14:46
The cap has been hit. It's all dependent on the accepts if it is to happen today. Certainly tomorrow! – Martijn Pieters Apr 02 '13 at 15:00
>>> Print Martjin Peters == 'Legend' >>> True Thank you! That's fantastic. So i'm now finding I can call things like "transaction2[72].date" This is great I was going to give each one a unique ID but perhaps this mitigates that need. Thanks! – Huw Apr 02 '13 at 15:41

Python- Automating Variable naming for with imported CSV data

1 Answers1