marcosan93

# Making a comparison DF
compare = pd.DataFrame()

compare['actual'] = y_test[col_name].reset_index(
    drop=True
)

compare['preds'] = preds

# Check to see if the predictions at least point in the correct direction

marcosan93

def transformData(df, days=1):
    """
    Transforming data into X variables for training.  Uses percent change and 
    multiplies the percentage by 100 rounded to 2 decimal places.
    """
    # Transforming data
    new_df = df.pct_change(
        days
    ).apply(
        lambda x: round(x*100, 2)

marcosan93

def getCryptoPrice(api_key, ticker="BTC-USD", n_days=1000):
    
    # Time periods
    now = datetime.now()

    # How far back to retrieve
    ago = now - timedelta(days=n_days)

    # Getting the price history for the crypto
    df = get_eod_data(

marcosan93

def matchMan(men_df, women_df, ratings, new_man_answers, num_sim=10):
    """
    This function will return the most likely compatible women based on a few given
    dataframes for a new male user.  Will use the top N similar users' compatibility 
    ratings to find the potentially most compatible women.
    """
    
    # First need to replace the DF answers with their numerical values
    men_df = men_df.apply(lambda x: x.cat.codes)

marcosan93

# Creating a Dataset of men and women
men = pd.DataFrame()

women = pd.DataFrame()

# Number of users
num = 1000

# Dating profile questions for each
qs = ['Q1', 'Q2', 'Q3', 'Q4', 'Q5']

marcosan93

def fbpBacktest(df, short=True):
    """
    Performs the final backtest using log returns and the positions function.
    Returns the performance.
    """
    # Getting positions
    df['positions'] = df.apply(lambda x: get_prophet_positions(x, short=short), axis=1)

    # Compensating for lookahead bias
    df['positions'] = df['positions'].shift(1)

marcosan93

def get_prophet_positions(df, short=True):
    """
    For these positions, buy when actual value is above the upper bound and short 
    when actual value is below lower bound. Otherwise do nothing.
    """
    if df['Open'] >= df['yhat_upper']:
        return 1
    elif df['Open'] <= df['yhat_lower'] and short:
        return -1
    else:

marcosan93

def runningFBP(ticker, window=730, ma_period=5, days_to_train=365, forecast_period=10):
    """
    Runs the facebook prophet model over the provided ticker.  Trains with last N days given 
    by days_to_train.  Forecast N days into the future based on given forecast_period.  Moving average 
    is applied to the dataset based on given ma_period. Returns the root mean squared error and a DF 
    of the actual values and the predicted values for the same day.
    """

    # Getting and Formatting Data
    df = getData(ticker, window=window, ma_period=ma_period)

marcosan93

def visFBP(df, forecast):
    """
    Given two dataframes: before training df and a forecast df, returns
    a visual chart of the predicted values and actual values.
    """
    # Visual DF
    vis_df = df[['ds','Open']].append(forecast).rename(
        columns={'yhat': 'Prediction',
                 'yhat_upper': "Predicted High",
                 'yhat_lower': "Predicted Low"}

marcosan93

def fbpTrainPredict(df, forecast_period):
    """
    Uses FB Prophet and fits to a appropriately formatted DF. Makes a prediction N days into 
    the future based on given forecast period. Returns predicted values as a DF.
    """
    # Setting up prophet
    m = Prophet(
        daily_seasonality=True, 
        yearly_seasonality=True, 
        weekly_seasonality=True