Problems running enamplecode on Apple Silicon (M1)

[sjachille]

Problem related to the use of nlpia on the new Apple Silicon M1 environment, using conda - installation is successful but there are many files that do not work

[sjachille]

Hello,
I have followed all the procedures I have been able to identify to install nlpia. While the installation is successful, there are many examples which do not work - see attached screenshots. Thank you for any help/suggestions you might have.

Code for example 4.1.5 follows so you can understand and hopefully help. The code is on Page 108, Ch. 4:

import pandas as pd
from nlpia.data.loaders import get_data
from sklearn.feature_extraction.text import TfidfVectorizer
from nltk.tokenize.casual import casual_tokenize
from sklearn.preprocessing import MinMaxScaler # See P.110
from nlpia.book.examples.ch04_catdog_lsa_3x6x16 import word_topic_vectors # from page 114

pd.options.display.width = 120
sms = get_data('sms-spam')
index = ['sms{}{}'.format(i, '!'*j) for (i, j) in zip(range(len(sms)), sms.spam)]

sms = pd.DataFrame(sms.values, columns=sms.columns, index=index)
sms['spam'] = sms.spam.astype(int)

print(len(sms))

print(sms.head(6))

P. 108 - Now let's do our tokenization and TF-IDF vector transformation on all these SMS messages:

tfidf_model = TfidfVectorizer(tokenizer=casual_tokenize)
tfidf_docs = tfidf_model.fit_transform(raw_documents=sms.text).toarray()

print(tfidf_docs.shape, sms.spam.sum())

P. 109

mask = sms.spam.astype(bool).values
spam_centeroid = tfidf_docs[mask].mean(axis=0)
ham_centeroid = tfidf_docs[~mask].mean(axis=0)

print(spam_centeroid.round(2))
print(ham_centeroid.round(2))

spamminess_score = tfidf_docs.dot(spam_centeroid - ham_centeroid)
print(spamminess_score.round(2))

P 110

sms['lda_score'] = MinMaxScaler().fit_transform(spamminess_score.reshape(-1, 1))
sms['lda_predict'] = (sms.lda_score > 0.5).astype(int)

print(sms['spam lda_predict lda_score'.split()].round(2).head(6))

Page 114 Listing 4.2

print(word_topic_vectors.T.round(1))