More Python Help

[paps511]

Hey All,

 

I have a .csv file with reviews for hotels. I need to take each of those reviews, and put them into the same string. Then convert it to a dictionary of words and the count of each word. I got it to work for a twitter assignment but now it keeps giving a 

TypeError                                 Traceback (most recent call last)
<ipython-input-1-dfe0abbe08b3> in <module>()
     13 for line in infile:
     14     print (line)
---> 15     text += line['text']
     16 
     17 words = text.split()

TypeError: string indices must be integers

 

I don't know where I am going wrong because it worked in one way and not the other. 

 

Here is my code

from collections import Counter
import csv

csvFile = open("FnLn-A2-Hotel1.csv", "a")
writer = csv.writer(csvFile)

text = ""
infile = open("HotelReviewSource.csv").readlines()
for line in infile:
#    print (line)
    text += line['text']

words = text.split()
cleantext = ""

for word in words:
    cleantext += word + " "
        
wordcount={}

for word in cleantext.split():
    if word not in wordcount:
        wordcount[word] = 1
    else:
        wordcount[word] += 1
print (wordcount)

c= Counter(cleantext.split())

for word,count in c.most_common():
    print (word,count)
    writer.writerow ([word,count])
    
csvFile.close()

 

[Enderman]

im not exactly sure what youre doing wrong because you didnt say what is going in and what you want out, but I know the problem is that "line[x]" returns the character which is in position x of the string "line"

for example if line is "hello" then line[2] will return "n"

 

so you cant have line['text'] because 'text' isnt a number, its a string

 

if you could give more info on what infile looks like and what you want lines 13-15 to output that would help

[Nineshadow]

10 minutes ago, Enderman said:

im not exactly sure what youre doing wrong because you didnt say what is going in and what you want out, but I know the problem is that "line[x]" returns the character which is in position x of the string "line"

for example if line is "hello" then line[2] will return "n"

 

so you cant have line['text'] because 'text' isnt a number, its a string

 

if you could give more info on what infile looks like and what you want lines 13-15 to output that would help

Unless line is a dictionary , but it isn't here and it doesn't make any sense for it to be a dictonary.

 

Anyway, you should probably do this :

for line in infile:
    text.append(line)

Where text is a buffer ( vector of strings )

 

Or

text+=line

If you just want text to be a big string, but I don't see the point since you use split anyway.

[paps511]

8 minutes ago, Enderman said:

im not exactly sure what youre doing wrong because you didnt say what is going in and what you want out, but I know the problem is that "line[x]" returns the character which is in position x of the string "line"

for example if line is "hello" then line[2] will return "n"

 

so you cant have line['text'] because 'text' isnt a number, its a string

 

if you could give more info on what infile looks like and what you want lines 13-15 to output that would help

Sorry, the input from the source file is a list reviews. In cell A1 is a review for a hotel down to about A65 etc

 

I need the final output to be a counter of words and the number of times they show up.

the 67

best 54

hotel 18

when 65

etc

 

After 'text' becomes just one big long sting of all the reviews, the rest of the code should work.

2 minutes ago, Nineshadow said:

Unless line is a dictionary , but it isn't here and it doesn't make any sense for it to be a dictonary.

 

Anyway, you should probably do this :

for line in infile:
    text+=line

 

HOT DAMN! It worked!!

That was a HUGE thorn in my ass.

 

I think I want to put you on retainer for the rest of my semester haha

[Nineshadow]

Just now, paps511 said:

Sorry, the input from the source file is a list reviews. In cell A1 is a review for a hotel down to about A65 etc

 

I need the final output to be a counter of words and the number of times they show up.

the 67

best 54

hotel 18

when 65

etc

 

After 'text' becomes just one big long sting of all the reviews, the rest of the code should work.

HOT DAMN! It worked!!

That was a HUGE thorn in my ass.

 

I think I want to put you on retainer for the rest of my semester haha

You know , errors are supposed to tell you what's wrong.

And here you had :

TypeError: string indices must be integers

At

 

text += line['text']

Which is pretty self-explanatory.

 

Anyway , good luck for the rest of your semester.

[paps511]

6 minutes ago, Nineshadow said:

You know , errors are supposed to tell you what's wrong.

And here you had :

TypeError: string indices must be integers

At

 

text += line['text']

Which is pretty self-explanatory.

 

Anyway , good luck for the rest of your semester.

The class is not really a coding class so much as it is data analysis. The professor just glosses over the coding aspect of it, but I'm making out a bit better than the rest which is not great..