Eligijus112

# Defining the dictionaries
cpu_dict = {
    'objective': 'reg:squarederror'
}

gpu_dict = {
    'objective': 'reg:squarederror',
    'tree_method': 'gpu_hist'
}

Eligijus112

# Defining the grid of parameters 
n_estimators = [30, 100, 200, 300, 600, 900]
max_depth = [4, 6, 8, 12, 14, 16]

# Number of rows to test on 
nrows = 30000

# Creating a dictionary
gpu_dict = {
    'objective': ['reg:squarederror'], 

Eligijus112

# Data wrangling 
import pandas as pd 

# Array math
import numpy as np 

# Quick value count calculator
from collections import Counter

Eligijus112

# Data wrangling 
import pandas as pd 

# Array math
import numpy as np 

# Quick value count calculator
from collections import Counter

Eligijus112

# self - an instance of the RandomForestClassifier class
# self.X - pandas dataframe containing feature information 
# self.Y - a list of binary response variable 
# self.X_obs_fraction - a float in range [0, 1] 

def bootstrap_sample(self):
    """
    Function that creates a bootstraped sample with the class instance parameters 
    """
    # Sampling the number of rows with repetition

Eligijus112

def best_split(self) -> tuple:
    """
    Given the X features and Y targets calculates the best split 
    for a decision tree
    """
    # Creating a dataset for spliting
    df = self.X.copy()
    df['Y'] = self.Y

    # Getting the GINI impurity for the base input 

Eligijus112

# Data wrangling 
import pandas as pd 

# Accuracy metrics 
from sklearn.metrics import precision_score, recall_score

# Reading data for classification 
d = pd.read_csv("data/random_forest/telecom_churn.csv")

# Setting the features used 

Eligijus112

# Data wrangling 
import pandas as pd 

# JSON functionalities
import json

# Reading a json in the .gz format 
def read_json(file_path:str) -> pd.DataFrame:
    """
    Reads the provided special JSON file format 

Eligijus112

# Importing all the methods for the pipeline 
from pipeline.modules.read_data import read_json
from pipeline.modules.clean_data import clean_text
from pipeline.modules.model_input_preparation import create_X_Y, apply_train_test_split
from pipeline.modules.model_fitting import TextCLF
from pipeline.modules.evaluate_model import eval_model

# Directory traversal 
import os 

Eligijus112

# Main frameworks to test out 
from pipeline.modules.read_data import read_json
from pipeline.modules.clean_data import clean_text

# Directory traversals 
import os 

def test_reading_cleaning():
    # Arrange 
    _path_to_data = os.path.join(