mberrien-fitzsimons

family_members = ['Misha', 'Kisha', 'Drew']
family_member.lower()

mberrien-fitzsimons

# read in libraries 
import os
import sys
import pandas as pd
import requests

from iexfinance.stocks import Stock 
from iexfinance.refdata import get_symbols

root_dir = os.path.join(os.getcwd(), '..')

mberrien-fitzsimons

# this blog of code allows me to import my secrets file from the conf folder in my 
# current jupyter notebook
root_dir = os.path.join(os.getcwd(), '..')
sys.path.append(root_dir)

from conf import secrets

# Download data as CSV from API 
function = 'TIME_SERIES_DAILY_ADJUSTED'
symbol = 'MSFT'

mberrien-fitzsimons

# import data from '01_raw' data folder into the processing notebook
msft_proc = pd.read_csv('../data/01_raw/msft_raw.csv', parse_dates=['timestamp'])

# Let's take a look at our dataset
msft_proc.head()

mberrien-fitzsimons

# Add days of the week 
msft_proc['day_of_week'] = msft_proc['timestamp'].dt.day_name()

# create dummy variables
dummies = pd.get_dummies(msft_proc['day_of_week'])

# drop original days of the week column from the original dataframe
msft_proc.drop(columns=['day_of_week'], inplace=True)

# add two dataframes together

mberrien-fitzsimons

# read new dataframe in from processed data folder
msft_model_df = pd.read_csv('../data/03_processed/msft_proc.csv')

mberrien-fitzsimons

# import required library 
from statsmodels.formula.api import ols

#Define the Problem
outcome = 'volume'
x_cols = ['Friday', 'Monday', 'Thursday', 'Tuesday', 'Wednesday']

#Fitting the actual model
predictors = '+'.join(x_cols)
formula = outcome + "~" + predictors

mberrien-fitzsimons

# save the model
filename = '../data/04_models/finalized_model.sav'
pickle.dump(model, open(filename, 'wb'))

mberrien-fitzsimons

# Put data collection code into a .py document in the src/d00_utils folder. From there 
# it can be imported into different jupyter notebooks for easy data 
def alphavantage_api_csv_download_raw(function, symbol, alpha_vantage_key):
    function = function
    symbol = symbol
    datatype = 'csv'
    url = f"https://www.alphavantage.co/query?function={function}&symbol={symbol}\
          &datatype={datatype}&apikey={ALPHA_VANTAGE_KEY}"
    return pd.read_csv(url)

mberrien-fitzsimons

# put data processing code into function
def process_alphavantage_data_create_dow_dummies(raw_data_file):
    raw_data_file['timestamp'] = pd.to_datetime(raw_data_file['timestamp'])
    raw_data_file['day_of_week'] = raw_data_file['timestamp'].dt.day_name()
    dummies = pd.get_dummies(raw_data_file['day_of_week'])
    raw_data_file.drop(columns=['day_of_week'], inplace=True)
    raw_data_file = pd.concat([raw_data_file, dummies], axis=1)
    # we are only interested in running a regression of volume against the dummy 
    # variables for days of the week. Because of this we will drop the remaining 
    # variables before importing it to our processed data folder