Home / Questions / 39090
Accepted
Eduardo Munizaga
Asked: 2020-12-14 17:45:52 +0800 CST 2020-12-14 17:45:52 +0800 CST

How to match between a dictionary and a list of lists in Python?

  • 772

What I need to do is score the words in a sentence that are closest to a predefined concept. What I have managed to do is exactly that but with only one sentence. I am attaching an example where if the answer phrase contains the word house, it is awarded 4 points, little house 3 and so on. If there are no matches it is zero and if there are more than one they are added. 

puntaje = {'casa':4, 'casita':3, 'casona':2, 'hogar':1}
respuesta = ['Esta', 'es', 'mi', 'casa', 'mi', 'hogar']

suma=[]

for k in respuesta:
        if k in puntaje:
              suma.append(puntaje[k])

print sum(suma)

In this example, it gives me 5 as a result , which is correct since in the sentence there are two matches with the dictionary: House that is worth 4 and Home that is worth 1 . Perfect.

My problem (apart from not knowing much about python) is that I can't do this with a list of multiple phrases (the idea is not to do another cycle, foralthough I couldn't either). I have been trying with compressed lists but I have not been able to find the result.

What I need is to parse a list like this:

respuestas = [['Esta', 'es', 'mi', 'casa', 'mi', 'hogar'], 
['Esta', 'es', 'una', 'casa', 'grande'],
['Esto', 'parece', 'una', 'casona'],
['Esto' 'es' 'un' 'departamento'],
[...etc]]

And the expected result would be the following:

{'Esta es mi casa mi hogar':5,
'Esta es una casa grande':4,
'Esto parece una casona':2,
'Esto es un departamento':0}
python

2 Answers

  1. Best Answer

    You will need two forbut it can be done in a single statement if you want the result to be that (a dictionary with the phrase as the key and the punctuation as the value). We simply use list comprehensions and the method joinfor strings:

    puntaje = {'casa':4, 'casita':3, 'casona':2, 'hogar':1}

respuestas = [['Esta', 'es', 'mi', 'casa', 'mi', 'hogar'],
['Esta', 'es', 'una', 'casa', 'grande'],
['Esto', 'parece', 'una', 'casona'],
['Esto', 'es', 'un', 'departamento']]

res={' '.join(frase):sum((puntaje[palabra]
    for palabra in frase if palabra in puntaje))
    for frase in respuestas}

print(res)

    Another option is to use the method count()to count the occurrences of each dictionary key in the lists:

    puntaje = {'casa':4, 'casita':3, 'casona':2, 'hogar':1}

respuestas = [['Esta', 'es', 'mi', 'casa', 'mi', 'hogar'],
['Esta', 'es', 'una', 'casa', 'grande'],
['Esto', 'parece', 'una', 'casona'],
['Esto', 'es', 'un', 'departamento']]

res={' '.join(frase):sum((frase.count(palabra)*puntos
        for palabra, puntos in puntaje.items()))
        for frase in respuestas}

print(res)

    Both ways can be done without using list comprehensions. The first code can be equivalent to:

    puntaje = {'casa':4, 'casita':3, 'casona':2, 'hogar':1}

respuestas = [['Esta', 'es', 'mi', 'casa', 'mi', 'hogar'],
['Esta', 'es', 'una', 'casa', 'grande'],
['Esto', 'parece', 'una', 'casona'],
['Esto', 'es', 'un', 'departamento']]

res = dict()
for frase in respuestas:
    aux = 0
    for palabra in frase:
        if palabra in puntaje:
            aux += puntaje[palabra]
    res[' '.join(frase)] = aux

print(res)

    And the second to:

    puntaje = {'casa':4, 'casita':3, 'casona':2, 'hogar':1}

respuestas = [['Esta', 'es', 'mi', 'casa', 'mi', 'hogar'],
['Esta', 'es', 'una', 'casa', 'grande'],
['Esto', 'parece', 'una', 'casona'],
['Esto', 'es', 'un', 'departamento']]

res = dict()
for frase in respuestas:
    aux = 0
    for palabra, puntos in puntaje.items():
        aux += frase.count(palabra)*puntos
    res[' '.join(frase)] = aux

print(res)

    And thanks to the 'magic' of Python in all four cases it comes out:

    {'This is an apartment': 0, 'This is a big house': 4, 'This is my house, my home': 5, 'This looks like a big house': 2}

    Note : All code is valid for Python 2.7 and Python3.x

    • 3

  2. A bit simpler than @FJSevilla's first answer using getthe dictionaries method:

    puntaje = {'casa':4, 'casita':3, 'casona':2, 'hogar':1}

respuestas = [
    ['Esta', 'es', 'mi', 'casa', 'mi', 'hogar'],
    ['Esta', 'es', 'una', 'casa', 'grande'],
    ['Esto', 'parece', 'una', 'casona'],
    ['Esto', 'es', 'un', 'departamento']
]

frases = (' '.join(palabras).lower() for palabras in respuestas)

res={ frase: sum(puntaje.get(palabra,0) for palabra in frase.split())
       for frase in frases }

print(res)
    • 2