from time import process_time, time
from sklearn.preprocessing import StandardScaler
import matplotlib.pyplot as plt
import tensorflow as tf
import utilities
import data
plt.style.use("seaborn-whitegrid")
# %% Generate data
# df = utilities.gen_ar_data()
# df = utilities.get_stock_data()
# df = utilities.get_weather_data()
df = data.get_energy_data()
df_train, df_test = utilities.split_ts(df)
df_train.plot()
df_test.plot()
forecast_gap = 170
train_len = 60
forecast_len = 24
# split into training samples
x_train, y_train = utilities.split_sequence(
df=df_train,
y_col="y",
train_len=train_len,
forecast_gap=forecast_gap,
forecast_len=forecast_len,
)
# %%
import numpy as np
import pandas as pd
from time import process_time, time
from sklearn.preprocessing import StandardScaler
import matplotlib.pyplot as plt
import utilities
import data
plt.style.use("seaborn-whitegrid")
# %% Generate data
df = data.get_weather_data()
df_train, df_test = utilities.split_ts(df)
df_train.plot()
df_test.plot()
# %% keras nbeats
forecast_gap = 100
train_len = 60
forecast_len = 24
# split into training samples
x_train, y_train = utilities.split_sequence(
df=df_train,
y_col="y",
train_len=train_len,
forecast_gap=forecast_gap,
forecast_len=forecast_len,
)