Creating a dictionary of dictionaries from csv file - string

Question

Hi so I am trying to write a function, classify(csv_file) that creates a default dictionary of dictionaries from a csv file. The first "column" (first item in each row) is the key for each entry in the dictionary and then second "column" (second item in each row) will contain the values.
However, I want to alter the values by calling on two functions (in this order):
trigram_c(string): that creates a default dictionary of trigram counts within the string (which are the values)
normal(tri_counts): that takes the output of trigram_c and normalises the counts (i.e converts the counts for each trigram into a number).
Thus, my final output will be a dictionary of dictionaries:
{value: {trigram1 : normalised_count, trigram2: normalised_count}, value2: {trigram1: normalised_count...}...} and so on
My current code looks like this:
def classify(csv_file):
l_rows = list(csv.reader(open(csv_file)))
classified = dict((l_rows[0], l_rows[1]) for rows in l_rows)
For example, if the csv file was:
Snippet1, "It was a dark stormy day"
Snippet2, "Hello world!"
Snippet3, "How are you?"
The final output would resemble:
{Snippet1: {'It ': 0.5352, 't w': 0.43232}, Snippet2: {'Hel' : 0.438724,...}...} and so on.
(Of course there would be more than just two trigram counts, and the numbers are just random for the purpose of the example).
Any help would be much appreciated!

Answer 1

First of all, please check classify function, because I can't run it. Here corrected version:
import csv
def classify(csv_file):
l_rows = list(csv.reader(open(csv_file)))
classified = dict((row[0], row[1]) for row in l_rows)
return classified
It returns dictionary with key from first column and value is string from second column.
So you should iterate every dictionary entry and pass its value to trigram_c function. I didn't understand how you calculated trigram counts, but for example if you just count the number of trigram appearence in string you could use the function below. If you want make other counting you just need to update code in the for loop.
def trigram_c(string):
trigram_dict = {}
start = 0
end = 3
for i in range(len(string)-2):
# you could implement your logic in this loop
trigram = string[start:end]
if trigram in trigram_dict.keys():
trigram_dict[trigram] += 1
else:
trigram_dict[trigram] = 1
start += 1
end += 1
return trigram_dict