crazytext: An Easy To Use Text Cleaning Package For NLP Built In Python
Some Times Text Can Become Very Crazy That The Content You Want and Really Useful Become Very Hard To Extract. crazytext is here to help you. It offers one line code snippets to clean and analyze your text faster than you.
why do the hard work when there is an option for smart work- Creator crazytext
pip install pandas
pip install numpy
pip install textblob
pip install sklearn
pip install lxml
pip install nltk
pip install crazytext
sample_text = 'AI is the future of HUMAN KIND, & Trendiest Topic of Today. #ai #future @aiforfuture https://ai.com (555) 555-1234 <p> Mobile Number </p> (555) 345-1234 <span>Pincode:</span> 224 'Let's Import Our Library
import crazytext as ct- Quick Analysis
doc = ct.Counter(text=sample_text)
doc.info()
>>
Length of String: 153
Number of URLs: 1
Number of Emails: 0
Number of Words: 25
Average Word Count: 6.12
Number of Stopwords: 4
Total Hashtags: 2
Total Mentions: 1
Total Length of Numeric Data: 7
Special Characters: 154
White Spaces: 28
Number of Vowels: 38
Number of Consonants: 143
Total Uppercase Words 3
Number of Phone Number Inside Text: 2
Observed Sentiment: (0.15, 'Positive')- Step By Step Analysis
doc.count_words()
>> 25
doc.count_stopwords()
>> 4
doc.count_phone_numbers()
>> 2
doc.count_uppercase_words()
>> 3You Can Try Many More Methods Just Type doc.count and press tab to get all the available Counter Methods.
Note : All The Methods For Counter Class Starts With count_
sample_text = 'AI is the future of HUMAN KIND, & Trendiest Topic of Today. #ai #future @aiforfuture www.ai.com (555) 555-1234 xyz@gmail.com <p> Mobile Number </p> (555) 345-1234 <span>Pincode:</span> 224 'Let's Import Our Library
import crazytext as ct
extractor = ct.Extractor(text=sample_text)Extracting Emails
extractor.get_emails()
>>['xyz@gmail.com']Extracting Phone Numbers
extractor.get_phone_numbers()
['(555) 555-1234', '(555) 345-1234']Extracting UPPER CASE words
extractor.get_uppercase_words()
>>['AI', 'HUMAN', 'KIND,']Extracting Hashtags
extractor.get_hashtags()
>>['#ai', '#future']Extracting Mentions
extractor.get_mentions()
>>['@aiforfuture']Extracting HTML Tags
extractor.get_html_tags()
>>['<p>', '</p>', '<span>', '</span>']Try Other Interesting Methods By Installing The Library Using pip install crazytext.
Note : All The Methods For Extractor Class Starts With get_
- There Are Two Ways To Clean The Text
- Remove Text Completly.
- Replace The Text With Its Saying
1. Remove Text Completly.
sample_text = '<h1>The Dark ó Knight</h1> a batman ó movie @batman ó #batman https://batman.com (555) 555-1234 ó 21 22 óó ó'Let's Import Our Library
import crazytext as ct
cleaner = ct.Cleaner(text=sample_text)Removing HTML Tags
cleaner.remove_html_tags_c()
>>' The Dark ó Knight a batman ó movie @batman ó #batman https://batman.com (555) 555-1234 ó 21 22 óó ó'Removing Phone Numbers
cleaner.remove_phone_numbers_c()
>> 'a batman ó movie @batman ó #batman https://batman.com ó 21 22 óó ó'2. Replace The Text With Its Saying Replacing HTML Tags
cleaner.remove_html_tags()
>>'HtmlTag The Dark ó Knight a batman ó movie @batman ó #batman https://batman.com (555) 555-1234 ó 21 22 óó ó'Replaxcing Phone Number
cleaner.remove_phone_numbers()
>> 'The Dark ó Knight</h1> a batman ó movie @batman ó #batman https://batman.com PhoneNumber ó 21 22 óó ó'To Clean A Doucment Quickly You Can Use quickclean() method inside Cleaner class.
Quick Clean
import crazytext as ct
ct = Cleaner(text=sample_text)
ct.quickclean(remove_complete=True,make_base=False)
>>'the dark knight batman movie batman batman'You Can Further Remove Duplicates Using The remove_duplicate_words() method.
Let's Load Hotel Reviews Dataframe From My Github.
import pandas as pd
df = pd.read_csv('https://raw.githubusercontent.com/Abhayparashar31/NLPP_sentiment-analsis-on-hotel-review/main/Restaurant_Reviews.tsv',delimiter = "\t",quoting=3)Let's Import Our Library and Creat A Object For Our Class Dataframe
import crazytext as ct
dc = ct.Dataframe(df=df,col='Review')Let's Find Our Dataframe Column Word Frequency Count Using crazytext
dc.get_df_words_frequency_count()
>>
the 405
and 378
I 294
was 292
a 228
...
Seat 1
dirty- 1
gross. 1
unbelievably 1
check. 1
Length: 2967, dtype: int64Cleaning The Dataframe Using One Line of Code With The Help of pretty text
df['cleaned_reviews'] = dc.clean(remove_complete=True,make_base='lemmatization')
df['cleaned_reviews']
>>
0 wow loved place
1 crust not good
2 not tasty texture nasty
3 stopped late may bank holiday rick steve recom...
4 the selection menu great price
....Next, Let's Convert This Cleaned Text Into Vectors For Further Processing
vector = ct.Dataframe(df=df,col='cleaned_reviews')
vector.to_tfidf(max_features=3500)
>>
array([[0. , 0. , 0. , 1. , 0. ],
[0. , 0.72888336, 0.6846379 , 0. , 0. ],
[0. , 0. , 1. , 0. , 0. ],
...,
[0. , 0. , 1. , 0. , 0. ],
[0. , 0. , 0. , 0. , 1. ],
[0. , 0. , 0. , 0. , 0. ]])- More NLP Tasks To Be Added.
- Inbuilt Model Support To Be Added.
We Are Unhappy To See You Go, You Can Give Your Feedback By Putting A Comment On The Repo.
pip uninstall crazytext