Develop Reference

Python dataframe to_cvs() how to remove the tab at the beginning of the output text file

I want to save a pandas DataFrame into a text file and make it R-friendly for later analysis. I used dataframe.to_cvs(filename, sep = '\t'). But I noticed that the output file begins with a tab, which is not quite readable for read.table() in R.
I used od -c filename, and it showed like this:
\t 1 2 3 4 \t 5 6 7 8 \t 1 2 3 ...
Is there any way to remove the tab at the beginning? Thank you in advance.

Looking at the documentation, it seems that this has to do with the index.
Try this one on for size:
dataframe.to_csv(filename, sep = '\t', index_label=False)
The docs are stating this:
index_labelstr or sequence, or False, default None
Column label for index column(s) if desired. If None is given, and header and index are True, then the index names are used. A sequence should be given if the object uses MultiIndex. If False do not print fields for index names. Use index_label=False for easier importing in R.
I am on Windows so I cannot use the od command to check.

Extract a string between other two in Python

I am trying to extract the comments from a fdf (PDF comment file). In practice, this is to extract a string between other two. I did the following:
I open the fdf file with the following command:
import re
import os
os.chdir("currentworkingdirectory")
archcom =open("comentarios.fdf", "r")
cadena = archcom.read()
With the opened file, I create a string called cadena with all the info I need. For example:
cadena = "\n215 0 obj\n<</D[2.0 2.0]/S/D>>\nendobj\n216 0 obj\n<</D[2.0 2.0]/S/D>>\nendobj\n217 0 obj\n<</D[2.0 2.0]/S/D>>\nendobj\n218 0 obj\n<</W 3.0>>\nendobj\n219 0 obj\n<</W 3.0>>\nendobj\ntrailer\n<</Root 1 0 R>>\n%%EOF\n"
I try to extract the needed info with the following line:
a = re.findall(r"nendobj(.*?)W 3\.0",cadena)
Trying to get:
a = "n216 0 obj\n<</D[2.0 2.0]/S/D>>\nendobj\n217 0 obj\n<</D[2.0 2.0]/S/D>>\nendobj\n218 0 obj\n<<"
But I got:
a = []
The problem is in the line a = re.findall(r"nendobj(.*?)W 3\.0",cadena) but I don't realize where. I have tried many combinations with no success.
I appreciate any comment.
Regards

It seems to me that there are 2 problems:
a) you are looking for nendobj, but the N is actually part of the line break \n. Thus you'll also not get a leading N in the output, because there is no N.
b) Since the text you're looking for crosses some newlines, you need the re.DOTALL flag
Final code:
a = re.findall("endobj(.*?)W 3\.0",cadena, re.DOTALL)
Also note, that there will be a second result, confirmed by Regex101.

How to specify a row delimiter in pandas read_csv()?

I would use a different row delimiter than \n with pandas.read_csv(). Does anyone know how to do that?
In the documentation of read_csv(), I found nothing related, but in the to_csv() page, I found the parameter line_terminator.
How may I solve this problem?

Here is an Example:
File: a.csv
hello,world:hell,worl:hel,wor:he,wo
a = pd.read_csv('a.csv', lineterminator=':')
print(a)
Output:
hello world
0 hell worl
1 hel wor
2 he wo

Data reading - csv

I have some datas in a .dfx file and I trying to read it as a csv with pandas. But it has some special characters which are not read by pandas. They are separators as well.I attached one line from it
The "DC4" is being removed when I print the file. The SI is read as space, correctly. I tried some encoding (utf-8, latin1 etc), but no success.
I attached the printed first line as well. I marked the place where the characters should be.
My code is simple:
import pandas
file_log = pandas.read_csv("file_log.DFX", header=None)
print(file_log)
I hope I was clear and someone has an idea.
Thanks in advance!
EDIT:
The input. LINK: drive.google.com/open?id=0BxMDhep-LHOIVGcybmsya2JVM28
The expected output:
88.4373 0 12.07.2014/17:05:22 38.0366 38.5179 1.3448 31.9839
30.0070 0 12.07.2014/17:14:27 38.0084 38.5091 0.0056 0.0033

By examining the example.DFX in hex (with xxd), the two separators are 0x14 and 0x0f accordingly.
Read the csv with multiple separators using python engine:
import pandas
sep1 = chr(0x14) # the one shows dc4
sep2 = chr(0x0f) # the one shows si
file_log = pandas.read_csv('example.DFX', header=None, sep='{}|{}'.format(sep1, sep2), engine='python')
print file_log
And you get:
0 1 2 3 4 5 6 7
0 88.4373 0 12.07.2014/17:05:22 38.0366 38.5179 1.3448 31.9839 NaN
1 30.0070 0 12.07.2014/17:14:27 38.0084 38.5091 0.0056 0.0033 NaN
It seems it has an empty column at the end. But I'm sure you can handle that.

The encoding seems to be ASCII here. DC4 stands for "device control 4" and SI for "Shift In". These are control characters in an ASCII file and not printable. Thus you cannot see them when you issue a "print(file_log)", although it might do something depending on your terminal to view this (like \n would do a new-line).
Try typing file_log in your interpreter to get the representation of that variable and check if those special characters are included. Chances are that you'll see DC4 in the representation as '\x14' which means hexadecimal 14.
You may then further process these strings in your program by using string manipulation like replace.

Python - Parsing Conundrum

I have searched high and low for a resolution to this situation, and tested a few different methods, but I haven't had any luck thus far. Basically, I have a file with data in the following format that I need to convert into a CSV:
(previously known as CyberWay Pte Ltd)
0 2019
01.com
0 1975
1 TRAVEL.COM
0 228
1&1 Internet
97 606
1&1 Internet AG
0 1347
1-800-HOSTING
0 8
1Velocity
0 28
1st Class Internet Solutions
0 375
2iC Systems
0 192
I've tried using re.sub and replacing the whitespace between the numbers on every other line with a comma, but haven't had any success so far. I admit that I normally parse from CSVs, so raw text has been a bit of a challenge for me. I would need to maintain the string formats that are above each respective set of numbers.
I'd prefer the CSV to be formatted as such:
foo bar
0,8
foo bar
0,9
foo bar
0,10
foo bar
0,11
There's about 50,000 entries, so manually editing this would take an obscene amount of time.
If anyone has any suggestions, I'd be most grateful.
Thank you very much.

If you just want to replace whitespace with comma, you can just do:
line = ','.join(line.split())
You'll have to do this only on every other line, but from your question it sounds like you already figured out how to work with every other line.

If I have correctly understood your requirement, you need a strip() on all lines and a split based on whitespace on even lines (lines starting from 1):
import re
fp = open("csv.txt", "r")
while True:
line = fp.readline()
if '' == line:
break
line = line.strip()
fields = re.split("\s+", fp.readline().strip())
print "\"%s\",%s,%s" % ( line, fields[0], fields[1] )
fp.close()
The output is a CSV (you might need to escape quotes if they occur in your input):
"Content of odd line",Number1,Number2
I do not understand the 'foo,bar' you place as header on your example's odd lines, though.