mikekeith52

import game
import pickle
import numpy as np

PCA_MODEL = pickle.load(open('model/pca.pckl','rb'))
LDA_MODEL = pickle.load(open('model/lda.pckl','rb'))
ReducedObsSpacePCA = PCA_MODEL.n_components_
ReducedObsSpaceLDA = len(LDA_MODEL.explained_variance_ratio_)

class State:

mikekeith52

    def evaluate_reward(self,game,action):
        init_points = game.points
        if ((len(game.dice) > 0) & (len([i for i in game.board if i.isnumeric()]) > 0)) | (game.jokers > 0):
            if game.board[action].isnumeric():
                next_die = game.dice[0]
                game.place_die(next_die,action+1)
            elif len([i for i in game.board if i.isnumeric()]) > 0:
                next_die = game.dice[0]
                adjusted_pos = min([int(i) for i in game.board if i.isnumeric()],key = lambda x: abs(x-(action+1)))
                game.place_die(next_die,adjusted_pos)

mikekeith52

    def _train(self):
        batch = np.asarray(random.sample(self.memory, BATCH_SIZE))
        if len(batch) < BATCH_SIZE:
            return

        current_states = []
        q_values = []
        max_q_values = []

        for entry in batch:

mikekeith52

expand_grid = lambda d: pd.DataFrame([row for row in product(*d.values())],columns=d.keys())
y = pd.Series(self.y)
y_train = y.values[:-test_length]
y_test = y.values[-test_length:]
dates = pd.to_datetime(self.current_dates)

scores = [] # lower is better
if y.min() > 0:
    grid = expand_grid({
        'trend':[None,'add','mul'],

mikekeith52

from Forecaster import Forecaster

forecasts = {}
for c in df.columns:
  f = Forecaster(y=df[c].to_list(),current_dates=df.index.to_list(),name=c)
  f.generate_future_dates(24,'MS') # MS is from pandas, meaning month start
  forecasts[c] = f

mikekeith52

from Forecaster import Forecaster

forecasts = {}
test_length = 3
for sym in ('UTPHCI','FLPHCI'):
    f = Forecaster()
    f.get_data_fred(sym)
    f.generate_future_dates(24,'MS')
    f.forecast_auto_arima(test_length=test_length)
    f.forecast_ets(test_length=test_length)

mikekeith52

def visualize_results():
    def on_button_clicked(b):
        ts_selection = ts_dropdown.value
        mo_selection = list(mo_select.value)
        with output:
            clear_output()
            forecasts[ts_selection].plot(models=mo_selection, print_mapes=True, plot_fitted=True)
      
    models = ('auto_arima','tbats','ets','average')
    ts_dropdown = widgets.Dropdown(options=forecasts.keys(), description = 'Time Series:')

mikekeith52

def accuracy_trend_vis():
    """ visualize the model error metrics
        leverages Jupyter widgets
    """
    def display_user_selections(ts_selection):
        """ displays graphs with seaborn based on what user selects from dropdown menus
        """
        f = forecasts[ts_selection]
        k = Counter(f.mape)
        mapes = [h[1] for h in k.most_common()]

mikekeith52

data_state.fillna(method='ffill',inplace=True)

max_perc = lambda x: np.max([i/100 for i in x])

data_state_piv = pd.pivot_table(data_state,
                                index='location',
                                values=['total_vaccinations_per_hundred','people_vaccinated','people_fully_vaccinated_per_hundred','people_fully_vaccinated'],
                                aggfunc={'total_vaccinations_per_hundred':max_perc,
                                       'people_fully_vaccinated_per_hundred':max_perc,
                                       'people_vaccinated':np.max,

mikekeith52

sns.set(rc={'figure.figsize':(14,8)})
ax3 = sns.lineplot(x=data_cumsum_doses.index,y='total_distributed',data=data_cumsum_doses,color='purple',linestyle='--',label='Total Distributed')
sns.lineplot(x=data_cumsum_doses.index,y='people_vaccinated',data=data_cumsum_doses,color='red',label='Received at Least One Dose')
sns.lineplot(x=data_cumsum_doses.index,y='people_fully_vaccinated',data=data_cumsum_doses,color='orange',label='Fully Vaccinated')

above_val = 5000000
plt.text(data_cumsum_doses.index.to_list()[-1], data_cumsum_doses['total_distributed'].max() + above_val, '{:,.0f}'.format(data_cumsum_doses['total_distributed'].max()),size=14,color = 'purple')
plt.text(data_cumsum_doses.index.to_list()[-1], data_cumsum_doses['people_vaccinated'].max() + above_val, '{:,.0f}'.format(data_cumsum_doses['people_vaccinated'].max()),size=14,color = 'red')
plt.text(data_cumsum_doses.index.to_list()[-1], data_cumsum_doses['people_fully_vaccinated'].max() + above_val, '{:,.0f}'.format(data_cumsum_doses['people_fully_vaccina