richiefrost

FROM ubuntu:latest

RUN apt-get update && \
    apt-get install python-dev libxml2-dev libxslt1-dev antiword unrtf poppler-utils pstotext tesseract-ocr \
flac ffmpeg lame libmad0 libsox-fmt-mp3 sox libjpeg-dev swig libpulse-dev python-pip -y && \
    pip install textract

richiefrost

def remove_null_cols(df):
    _df = df.copy()
    _df = df.dropna(how='all', axis=1)
    return  _df

def set_category_types(df, columns):
    _df  = df.copy()
    for col in columns:
        _df[col]  = df[col].astype('category')
    return _df

richiefrost

from azure.datalake.store import core, lib, multithread
import pandas as pd

class ADLSHelper:
    def __init__(self, store_name='mystorename'):
        """
        When initializing this helper, it will prompt you to do an interactive login to connect to your data lake store.
        It uses Azure Active Directory for authentication, and you use the token returned from
        your login process to connect to your Azure Data Lake instance.
        You can also authenticate with username/password or ServicePrincipal for production.

richiefrost

from time import time

def log_time(func):
    """Logs the time it took for func to execute"""
    def wrapper(*args, **kwargs):
        start = time()
        val = func(*args, **kwargs)
        end = time()
        duration = end - start
        print(f'{func.__name__} took {duration} seconds to run')

richiefrost

class QueryBuilder:
     def __init__(self):
          self.select_value = ''
          self.from_table_name = ''
          self.where_value = ''
          self.groupby_value = ''

     def select(self, select_arg):
          self.select_value = select_arg
          return self

richiefrost

# Don't do this
def get_data_bad(query_text):
    db = SQLDB()
    return db.get(query_text)
    
# What if you need to use a DocDB instance? Or a DynamoDB instance?
# Do this instead
def get_data(db, query_text):
    return db.get(query_text)
    

richiefrost

import pandas as pd
import numpy as np
from sklearn.cluster import KMeans

df = pd.read_csv('user_history.csv')

# Pandas' cut method groups continuous values into equal-sized bins
df['Frequency'] = pd.cut(df['RequestsPerMonth'], bins=4)

# Since lower recency is better, we need to reverse the order of the bins

richiefrost

from sklearn.tree import DecisionTreeClassifier
import pandas as pd

# Get the most valuable customers, from step 2
df = pd.read_csv('high_value_customers.csv')

# Churned is our target. Why did they/didn't they churn?
X, y = df.drop('Churned', axis=1), df['Churned']
model = DecisionTreeClassifier()
model.fit(X, y)