How to calculate a distance in python on 2 words

user_11334913 0
Viewed 45

I have to extract a database and put it in a script in python. I did it for all name of database. Now I have to calculate the difference of letter between the name 1 and the name 2 and between the name 1 and the name 3 .... and between the name 2 with the name 3 .....

I did it for the moment. I show you what a little part about what I did who gives me a problem :

valeur = []

c.execute("""SELECT  NAME FROM AUTHORS""")
rows = c.fetchall()
for row in rows:
    valeur.append(row)
x=0
liste=[]
for x in range(len(valeur)):
    for i in range(x+1,len(valeur)):
        ed = nltk.edit_distance(valeur[x],valeur[i])
        dico={"titre":valeur[x],"titre_compare":valeur[i], "distance":ed}
        dico["titre"],"&&",(dico["titre_compare"]),"&&",(dico["distance"])
        print(dico["titre"],"&&",(dico["titre_compare"]),"&&",(dico["distance"]))

And the result is here


('Khālid Muḥammad ʻAlī al-Ḥājj',) && ('K. Hardono',) && 1
('Khālid Muḥammad ʻAlī al-Ḥājj',) && ('Nūr al-Dīn Ṣammūd',) && 1
('Khālid Muḥammad ʻAlī al-Ḥājj',) && ('Mawlūd Qāsim Nāyit Bulqāsim',) && 1
('Khālid Muḥammad ʻAlī al-Ḥājj',) && ('Ali Ak',) && 1
('Khālid Muḥammad ʻAlī al-Ḥājj',) && ('Nazamī Rāmajī',) && 1

My problems are why I have only 1 at the end of the comparaison between the 2 words and why I have a "," at the end of the words, It should be different and I don't unterstand why I got this. I did the same code of a different database with tittle of books and it worked.

Thanks you very much if you can help me !

nltk python database
3 Answers

why I have a "," at the end of the words

You have tuples that contain a single string, instead of just having the strings. The reason you have these is because you got them from a database query, which gives you each row in a tuple, no matter how many columns there are.

why I have only 1 at the end of the comparaison between the 2 words

This appears to have the same cause: nltk is being given the wrong sort of data. It would compare strings character by character; it compares tuples element by element. There is one element in each tuple, and they are different between the two tuples, so of course the result is 1.

user_5236120

your issues are caused by the fact cursor.fetchall() returns all the rows as a list of tuples, to solve this you can use:

valeur = [r[0] for r in rows]

@KarlKnechtel already explained why you get your output in that form

user_31615750

Thanks for your answer, So I dit it :

valeur=[]

req = "select * from AUTHORS"
result = c.execute(req)
print(type(result))
for row in result:
    #print(row[1])
    #print(row[2])
    valeur.append(row[0])

x=0
liste=[]
for x in range(len(valeur)):
    for i in range(x+1,len(valeur)):
        ed = nltk.edit_distance(valeur[x],valeur[i])
        dico={"titre":valeur[x],"titre_compare":valeur[i], "distance":ed}
        (dico["titre"],"&&",(dico["titre_compare"]),"&&",(dico["distance"]))
        double=dico["distance"]+dico["distance"]
        print(dico["titre"],"&&",(dico["titre_compare"]),"&&",(double))

and it works perfectely but I have not the name, I have it :

/authors/OL1022167A && /authors/OL1022282A && 6
/authors/OL1022167A && /authors/OL102236A && 4
/authors/OL1022167A && /authors/OL102364A && 6
/authors/OL1022167A && /authors/OL1023681A && 8

But me I want the name and here it's not the name. SO I tried to say
"select name from authors"

But when i said it , it doesn't work. So I tried other things but it always doesn't work. I would like to know How I can do like I did but only for name and not for key.

Thanks again and thanks for your previous answer !!!!

user_113349130
Related