sksoumik

# import necessary libraries 

import pandas as pd 
import matplotlib.pyplot as plt
import numpy as np
import warnings

# ignore warnings
warnings.filterwarnings("ignore")

sksoumik

from google.colab import drive
drive.mount('/content/drive')

sksoumik

data['split'] = np.random.randn(data.shape[0], 1)

msk = np.random.rand(len(data)) <= 0.7

train = data[msk]
test = data[~msk]

train = train.filter(items=['comment', 'class_id'])
test = test.filter(items=["comment", "class_id"])

sksoumik

# make sure colab runs tf 2.0
try:
    %tensorflow_version 2.x
except Exception:
    pass
import tensorflow as tf

sksoumik

def confusion_matrix_data(true_label:int, predicted_label:int):
    data = (label_df
                .loc[lambda d: d['Pred'] == predicted_label]
                .loc[lambda d: d['class_id'] == true_label]
                ['comment']) 
    return data

false_positive = confusion_matrix_data(0, 1)
print(false_positive[:3])

sksoumik

from sklearn.model_selection import train_test_split
from sklearn.preprocessing import LabelEncoder,OneHotEncoder


'''
convert target column texts into numeric values
'''
# convert text sentiments into numeric values
sentiment_to_id = {
    "empty": 0,

sksoumik

# Basic few methods to clean text data
"""
Install the following libraries
!pip install tweet-preprocessor
!pip install emoji 

"""


"""

sksoumik

import preprocessor as p 
# set opt(option) as MENTION and URL
p.set_options(p.OPT.MENTION, p.OPT.URL)
p.clean("hello guys @alx #sport🔥 1245 https://github.com/s/preprocessor")
# apply to a specfic column and add a new column with the new clean data. 
data['clean_content'] = data.content.apply(lambda x: p.clean(x))

sksoumik

import warnings
import pandas as pd

# ignore warnigs
warnings.filterwarnings("ignore")

# force output to display the full description of pandas 
pd.set_option('display.max_colwidth', -1)

df = pd.read_csv('PATH')

sksoumik

# sample_dict is an dict()
# print 3 key and value from the dictionay. 
N = 3
{key:value for key,value in list(sample_dict.items())[0:N]}