import numpy as np
import pandas as pd
import plotly.offline
from statsmodels.tsa.arima_model import ARMA
from statsmodels.tsa.ar_model import AR
from sklearn.metrics import mean_squared_error as mse
import plotly.plotly as py
import plotly.graph_objs as go
import warnings




# simple moving average
def SMA(deg, threshold, data) -> tuple:
    packets_sent = deg
    data_copy = data[:deg]
    warnings.filterwarnings('ignore')
    for i in range(deg, len(data)):
        model = ARMA(data_copy[i - deg:], order=(0, 0))
        model = model.fit(disp=False)
        pred = model.predict(0, 0)
        if abs(data[i] - pred) >= threshold:
            data_copy = np.append(data_copy, [data[i]])
            packets_sent += 1
        else:
            data_copy = np.append(data_copy, [pred])
    return packets_sent, mse(data, data_copy)


# auto regressive model
def ARModel(deg, threshold, data) -> tuple:
    packets_sent = deg + 2
    data_copy = data[:deg + 2]
    for i in range(deg + 2, len(data)):
        model = AR(data_copy)
        model = model.fit(disp=False, maxlag=deg)
        pred = model.predict(i, i)
        if abs(data[i] - pred) >= threshold:
            data_copy = np.append(data_copy, [data[i]])
            packets_sent += 1
        else:
            data_copy = np.append(data_copy, pred)
    return packets_sent, mse(data, data_copy)


# calculate packages sent and MSE with different algorithms at different frequencies(sampling)
# Eg. frequency = 4 means we take every 4 data point(slower sampling)
def calculations(data, maxThreshold, step, data_type, algos_list, plot=True, freq=1):
    if freq > 1:
        data = data[::freq]
    packets_sent_thresh2 = []
    packets_sent_thresh4 = []
    packets_sent_ar2 = []
    packets_sent_ar3 = []
    mse_err2 = []
    mse_err4 = []
    mse_err_ar2 = []
    mse_err_ar3 = []
    for thresh in range(0, maxThreshold + 1, step):
        ps, er = SMA(2, thresh, data)
        packets_sent_thresh2.append((ps / len(data)) * 100)
        mse_err2.append(er)
        ps, er = SMA(4, thresh, data)
        packets_sent_thresh4.append((ps / len(data)) * 100)
        mse_err4.append(er)
        ps, er = ARModel(2, thresh, temperatures)
        packets_sent_ar2.append((ps / len(data)) * 100)
        mse_err_ar2.append(er)
        ps, er = ARModel(3, thresh, data)
        packets_sent_ar3.append((ps / len(data)) * 100)
        mse_err_ar3.append(er)

    vals = [packets_sent_thresh2, packets_sent_thresh4, packets_sent_ar2, packets_sent_ar3]
    err = [mse_err2, mse_err4, mse_err_ar2, mse_err_ar3]

    if plot:
        x = np.array(range(0, maxThreshold + 1, step))
        ps_algo_comparison(x, vals, data_type, algos_list)
        err_algo_comparison(x, err, data_type, algos_list)

    return vals, err


# compare packages sent percentage over different algorithms
def ps_algo_comparison(x, y_list, data_type, algos_list):
    data = []
    for i in range(len(y_list)):
        obj = go.Scatter(x=x, y=y_list[i], name=algos_list[i], showlegend=True)
        data.append(obj)
    layout = go.Layout(title='Packages Sent/Threshold ratio for '+data_type, xaxis=dict(title='Threshold'), yaxis=dict(title='% of sent packages'))
    fig = go.Figure(data, layout)
    plotly.offline.plot(fig, filename='ps-algorithms')


# compare MSE over different algorithms
def err_algo_comparison(x, errs, data_type, algos_list):
    data = []
    for i in range(len(errs)):
        obj = go.Scatter(x=x, y=errs[i], name=algos_list[i], showlegend=True)
        data.append(obj)
    layout = go.Layout(title='MSE/Threshold ratio for '+data_type, xaxis=dict(title='Threshold'), yaxis=dict(title='MSE'))
    fig = go.Figure(data, layout)
    plotly.offline.plot(fig, filename='mse-algorithms')


# compare packages sent percentage over data sampled at different frequencies for certain algorithm
def ps_plot_frequencies(x, algo, y_list, freq, data_type):
    data = []
    for i, y in enumerate(y_list):
        obj = go.Scatter(x=x, y=y, name="Normal Frequency" if freq[i] == 1 else f'Frequency({freq[i]})', showlegend=True)
        data.append(obj)
    layout = go.Layout(title='Packages Sent/Threshold with different data frequencies for ' + data_type+' with '+algo, xaxis=dict(title='Threshold'),
                       yaxis=dict(title='% of sent packages'))
    fig = go.Figure(data, layout)
    plotly.offline.plot(fig, filename=algo+' frequency for '+data_type)


def calc_freq(data, maxThreshold, step, algorithms_list, data_header):
    tt = []
    freq = [1, 2, 4]
    for f in freq:
        ps_algos, err_algos = calculations(data, maxThreshold, step, algos_list=algorithms_list, data_type=data_header, freq=f, plot=False)
        tt.append(ps_algos)

    for algorithm, rez in zip(algs, zip(*tt)):
        ps_plot_frequencies(np.array(range(0, maxThresh_temp + 1, temp_step)), algorithm, rez, freq, columns[1])


# getting the dataset and preprocess it
df = pd.read_csv("April.csv")
df = df.loc[df["Station Name"] == "Stripa Granite)"]
df = df[["Data Time", "Room Temperature °C"]]
df["Data Time"] = pd.to_datetime(df["Data Time"])
df = df.loc[(df["Data Time"] >= '5/28/2014  8:31:00') & (df["Data Time"] <= '6/21/2014  3:49:00')]
df.sort_values(by=["Data Time"], inplace=True, ascending=True)
temperatures = df["Room Temperature °C"].values
columns = df.columns.tolist()

# plot the timeseries for temperature and wind speed
temperature = go.Scatter(x=df["Data Time"], y=temperatures, name='Room Temperature °C('+'\u2103'+')', showlegend=True)
data = [temperature]
layout = go.Layout(title=columns[1]+"/"+columns[2], xaxis=dict(title='Date'))
fig = go.Figure(data=data, layout=layout)
plotly.offline.plot(fig, filename='time-series-temperature')

maxThresh_temp = 8
temp_step = 1


algs = ['SMA(2)', 'SMA(4)', 'AR(2)', 'AR(3)']