#!/usr/bin/env python
# coding: utf-8

# In[1]:


import pandas as pd
from collections import defaultdict
import warnings

warnings.filterwarnings("ignore")

import pandas as pd
import plotly

import greykite
from greykite.common.data_loader import DataLoader
from greykite.framework.templates.autogen.forecast_config import ForecastConfig, ComputationParam, ModelComponentsParam
from greykite.framework.templates.autogen.forecast_config import MetadataParam
from greykite.framework.templates.forecaster import Forecaster
from greykite.framework.templates.model_templates import ModelTemplateEnum
from greykite.framework.utils.result_summary import summarize_grid_search_results

from fbprophet import Prophet


# ## peyton_manning

# In[2]:


# Loads dataset into pandas DataFrame
dl = DataLoader()
df = dl.load_peyton_manning()


# In[3]:


# specify dataset information
metadata = MetadataParam(
    time_col='ts',  # name of the time column ("date" in example above)
    value_col='y',  # name of the value column ("sessions" in example above)
    freq='D'# "H" for hourly, "D" for daily, "W" for weekly, etc.
                # Any format accepted by `pandas.date_range`
)


# ### SILVERKITE

# In[4]:


forecaster = Forecaster()  # Creates forecasts and stores the result
result = forecaster.run_forecast_config(  # result is also stored as `forecaster.forecast_result`.
    df=df,
    config=ForecastConfig(
        model_template=ModelTemplateEnum.SILVERKITE.name,
        forecast_horizon=365,  # forecasts N steps ahead
        coverage=0.95,         # 95% prediction intervals
        metadata_param=metadata
    )
)


# In[5]:


backtest = result.backtest
fig = backtest.plot()
plotly.io.show(fig)


# In[6]:


forecast = result.forecast
fig = forecast.plot_components()
plotly.io.show(fig)     # fig.show() if you are using "PROPHET" template


# In[7]:


grid_search = result.grid_search
cv_results = summarize_grid_search_results(
    grid_search=grid_search,
    decimals=2,
    # The below saves space in the printed output. Remove to show all available metrics and columns.
    cv_report_metrics=None,
    column_order=["rank", "mean_test", "split_test", "mean_train", "split_train", "mean_fit_time", "mean_score_time", "params"]
)
# Transposes to save space in the printed output
cv_results["params"] = cv_results["params"].astype(str)
cv_results.set_index("params", drop=True, inplace=True)
cv_results.transpose()


# ### SILVERKITE(gradient_boosting)

# In[8]:


forecaster = Forecaster()  # Creates forecasts and stores the result
result = forecaster.run_forecast_config(  # result is also stored as `forecaster.forecast_result`.
    df=df,
    config=ForecastConfig(
        model_template=ModelTemplateEnum.SILVERKITE.name,
        forecast_horizon=365,  # forecasts N steps ahead
        coverage=0.95,         # 95% prediction intervals
        metadata_param=metadata,
        model_components_param=ModelComponentsParam(
            custom=dict(fit_algorithm_dict=dict(fit_algorithm="gradient_boosting"))
        )
    )
)


# In[9]:


backtest = result.backtest
fig = backtest.plot()
plotly.io.show(fig)


# In[10]:


grid_search = result.grid_search
cv_results = summarize_grid_search_results(
    grid_search=grid_search,
    decimals=2,
    # The below saves space in the printed output. Remove to show all available metrics and columns.
    cv_report_metrics=None,
    column_order=["rank", "mean_test", "split_test", "mean_train", "split_train", "mean_fit_time", "mean_score_time", "params"]
)
# Transposes to save space in the printed output
cv_results["params"] = cv_results["params"].astype(str)
cv_results.set_index("params", drop=True, inplace=True)
cv_results.transpose()


# ### PROPHET

# In[11]:


forecaster = Forecaster()  # Creates forecasts and stores the result
result = forecaster.run_forecast_config(  # result is also stored as `forecaster.forecast_result`.
    df=df,
    config=ForecastConfig(
        model_template=ModelTemplateEnum.PROPHET.name,
        forecast_horizon=365,  # forecasts N steps ahead
        coverage=0.95,         # 95% prediction intervals
        metadata_param=metadata,
    )
)


# In[12]:


backtest = result.backtest
fig = backtest.plot()
plotly.io.show(fig)


# In[13]:


forecast = result.forecast
fig = forecast.plot_components()
fig.show()


# In[14]:


grid_search = result.grid_search
cv_results = summarize_grid_search_results(
    grid_search=grid_search,
    decimals=2,
    # The below saves space in the printed output. Remove to show all available metrics and columns.
    cv_report_metrics=None,
    column_order=["rank", "mean_test", "split_test", "mean_train", "split_train", "mean_fit_time", "mean_score_time", "params"]
)
# Transposes to save space in the printed output
cv_results["params"] = cv_results["params"].astype(str)
cv_results.set_index("params", drop=True, inplace=True)
cv_results.transpose()


# ## AirPassengers

# In[15]:


df = pd.read_csv('../data/AirPassengers.csv')
df['Month'] = pd.to_datetime(df['Month'])


# In[16]:


# specify dataset information
metadata = MetadataParam(
    time_col="Month",  # name of the time column ("date" in example above)
    value_col="#Passengers",  # name of the value column ("sessions" in example above)
    freq="MS"# "H" for hourly, "D" for daily, "W" for weekly, etc.
                # Any format accepted by `pandas.date_range`
)


# ### SILVERKITE

# In[17]:


forecaster = Forecaster()  # Creates forecasts and stores the result
result = forecaster.run_forecast_config(  # result is also stored as `forecaster.forecast_result`.
    df=df,
    config=ForecastConfig(
        model_template=ModelTemplateEnum.SILVERKITE.name,
        forecast_horizon=12,  # forecasts N steps ahead
        coverage=0.95,         # 95% prediction intervals
        metadata_param=metadata
    )
)


# In[18]:


backtest = result.backtest
fig = backtest.plot()
plotly.io.show(fig)


# In[19]:


forecast = result.forecast
fig = forecast.plot_components()
plotly.io.show(fig)     # fig.show() if you are using "PROPHET" template


# In[20]:


grid_search = result.grid_search
cv_results = summarize_grid_search_results(
 grid_search=grid_search,
 decimals=2,
 # The below saves space in the printed output. Remove to show all available metrics and columns.
 cv_report_metrics=None,
 column_order=["rank", "mean_test", "split_test", "mean_train", "split_train", "mean_fit_time", "mean_score_time", "params"])
# Transposes to save space in the printed output
cv_results["params"] = cv_results["params"].astype(str)
cv_results.set_index("params", drop=True, inplace=True)
cv_results.transpose()


# ### SILVERKITE(gradient_boosting)

# In[21]:


forecaster = Forecaster()  # Creates forecasts and stores the result
result = forecaster.run_forecast_config(  # result is also stored as `forecaster.forecast_result`.
    df=df,
    config=ForecastConfig(
        model_template=ModelTemplateEnum.SILVERKITE.name,
        forecast_horizon=12,  # forecasts N steps ahead
        coverage=0.95,         # 95% prediction intervals
        metadata_param=metadata,
        model_components_param=ModelComponentsParam(
            custom=dict(fit_algorithm_dict=dict(fit_algorithm='gradient_boosting'))
        )
    )
)


# In[22]:


backtest = result.backtest
fig = backtest.plot()
plotly.io.show(fig)


# In[23]:


grid_search = result.grid_search
cv_results = summarize_grid_search_results(
 grid_search=grid_search,
 decimals=2,
 # The below saves space in the printed output. Remove to show all available metrics and columns.
 cv_report_metrics=None,
 column_order=["rank", "mean_test", "split_test", "mean_train", "split_train", "mean_fit_time", "mean_score_time", "params"])
# Transposes to save space in the printed output
cv_results["params"] = cv_results["params"].astype(str)
cv_results.set_index("params", drop=True, inplace=True)
cv_results.transpose()


# ### PROPHET

# In[24]:


forecaster = Forecaster()  # Creates forecasts and stores the result
result = forecaster.run_forecast_config(  # result is also stored as `forecaster.forecast_result`.
 df=df,
 config=ForecastConfig(
     model_template=ModelTemplateEnum.PROPHET.name,
     forecast_horizon=12,  # forecasts 365 steps ahead
     coverage=0.95,         # 95% prediction intervals
     metadata_param=metadata,
     model_components_param=ModelComponentsParam(
         seasonality=dict(seasonality_mode='multiplicative')
     )
 )
)


# In[25]:


backtest = result.backtest
fig = backtest.plot()
plotly.io.show(fig)


# In[26]:


forecast = result.forecast
fig = forecast.plot_components()
#plotly.io.show(fig)     # fig.show() if you are using "PROPHET" template
fig.show()


# In[27]:


grid_search = result.grid_search
cv_results = summarize_grid_search_results(
 grid_search=grid_search,
 decimals=2,
 # The below saves space in the printed output. Remove to show all available metrics and columns.
 cv_report_metrics=None,
 column_order=["rank", "mean_test", "split_test", "mean_train", "split_train", "mean_fit_time", "mean_score_time", "params"])
# Transposes to save space in the printed output
cv_results["params"] = cv_results["params"].astype(str)
cv_results.set_index("params", drop=True, inplace=True)
cv_results.transpose()