David-Woroniuk

Argument	        Description

api_key	          the API key obtained from https://www.worldweatheronline.com/developer/. (str)
attribute_list	  a list of weather attributes to collect. (list)
city	            a city for which to retrieve data. (str).
start_date	      a string in the format YYYY-MM-DD (str).
end_date	        a string in the format YYYY-MM-DD (str).
frequency	        the frequency of extracted data, measured in hours. (int)
verbose	          boolean determining printing during data extraction. (bool) [Default = True]
csv_directory	    an optional file directory to store the output. (os directory) [Default = None]

David-Woroniuk

dataset = RetrieveByAttribute(api_key, attribute_list, city, start_date, end_date, frequency).retrieve_hist_data()

David-Woroniuk

dataset = HistoricalLocationWeather(api_key, city, start_date, end_date, frequency).retrieve_hist_data()

David-Woroniuk

attributes = DetermineListOfAttributes(api_key, verbose = True).retrieve_list_of_options()

David-Woroniuk

pip install WorldWeatherPy

from WorldWeatherPy import DetermineListOfAttributes
from WorldWeatherPy import HistoricalLocationWeather
from WorldWeatherPy import RetrieveByAttribute

David-Woroniuk

def plot_outliers(data):
  '''
  This function determines the position of the outliers within the time-series,
  which are subsequently plotted.

  Arguments: 
          data: the initial dataset (Pandas DataFrame).

  Returns:
          fig: a visual representation of the outliers present in the series, as

David-Woroniuk

def generate_testing_loss(test_value):
  '''
  This function uses the model to predict anomalies within the test set.
  Additionally, this function generates the 'anomalies' global variable,
  containing the outliers identified by the RNN.

  Arguments: 
          test_value: an array of testing values (numpy array).

  Returns:

David-Woroniuk

# now call the above function:
test_value = normalise_testing_values(testing_data['Volume'], training_mean, training_std) 

David-Woroniuk

def normalise_testing_values(data, training_mean, training_std):
  '''
  This function uses the training mean and standard deviation to normalise
  the testing data, generating a numpy array of test values.

  Arguments: 
          data: the data to be used in model testing (Pandas DataFrame column).
          mean: the training set mean (float).
          std: the training set standard deviation (float).

David-Woroniuk

def reconstruction_error(x_train):
  '''
  This function calculates the reconstruction error and displays a histogram of
  the training mean absolute error.

  Arguments: 
          x_train: 3D data to be used in model training (numpy array).

  Returns:
          fig: a visual representation of the training MAE distribution.