import pandas as pd from prophet import Prophet

df = pd.read_csv('uk_cases_by_publish_date_06122021.csv') df.head()

df = df[["date", "newCasesByPublishDate"]]

df = df.rename(columns={"newCasesByPublishDate": "y"})

df.info()

df["date"] = pd.to_datetime(df["date"], format = "%d/%m/%Y")

df = df.rename(columns={"date": "ds"})

#df['floor'] = 0 #df['cap'] = 100000 m = Prophet(yearly_seasonality=True, seasonality_mode='multiplicative') m.add_country_holidays(country_name='UK') m.fit(df)

future = m.make_future_dataframe(periods=365) #future['floor'] = 0 #future['cap'] = 100000 future.tail()

forecast = m.predict(future) forecast[['ds', 'yhat', 'yhat_lower', 'yhat_upper']].tail()

# Python fig1 = m.plot(forecast)

# Python fig2 = m.plot_components(forecast)

from prophet.plot import plot_plotly, plot_components_plotly plot_plotly(m, forecast)

plot_components_plotly(m, forecast)

forecast = m.predict(future) fig = m.plot_components(forecast)

forecast.head()

forecast.to_csv("UK Covid Cases Predictions.csv", index=False)

from sklearn.metrics import mean_absolute_percentage_error

forecast_pred = forecast[["ds", "yhat", "yhat_lower", "yhat_upper"]]

df.head()

df_combined = df.merge(forecast_pred, on="ds", how="left")

df_combined.head()

mean_absolute_percentage_error(df_combined["y"], df_combined["yhat"])

from sklearn.metrics import mean_absolute_error

mean_absolute_error(df_combined["y"], df_combined["yhat"])

df2021 = df_combined[(df_combined['ds'] > '2021-01-01') & (df_combined['ds'] < '2021-12-06')]

mean_absolute_percentage_error(df2021["y"], df2021["yhat"])

df2021.to_csv("2021_UK_COVID_finalpredictions.csv", index=False)