Cosine Similarity with tfidf using Pyspark dataframe

Question

I am able to generate the Cosine Similarity using Pandas dataframe. But, I need to do it on Pyspark using spark dataframe. I didn't find any relevant info over google and I am new with Pyspark so reaching out.

corpus = [
    "Hello there good man!",
    "It is quite windy in ",
    "How is the weather today?",
    "How isssss ay?",
    "windy London is where"
]

query = "windy London"

import pandas as pd
from sklearn.feature_extraction.text import TfidfVectorizer

tfidf_vectorizer = TfidfVectorizer()
tfidf_matrix = tfidf_vectorizer.fit_transform(corpus)
print(tfidf_matrix.shape)


#Optional to see the data
doc_term_matrix = tfidf_matrix.todense()
df = pd.DataFrame(doc_term_matrix, 
                  columns=tfidf_vectorizer.get_feature_names())


#Cosine Similarity
from sklearn.metrics.pairwise import cosine_similarity

queryTFIDF = TfidfVectorizer().fit(corpus)

queryTFIDF = queryTFIDF.transform([query])

cosine_similarities = cosine_similarity(queryTFIDF, tfidf_matrix).flatten()

related_product_indices = cosine_similarities.argsort()[-1] #For first prediction