[Practice] Predict temperature

weather.csv

weather.py

#%%
import pandas as pd

weather = pd.read_csv('weather.csv', sep=',', encoding='utf-8')
train = list(weather[(weather['年'] <= 2017)]['気温'])
test = list(weather[(weather['年'] >= 2018)]['気温'])

#%%
interval = 6
def generate_x_y(data):
    x = []
    y = []

    for i in range(len(data)):
        if i < interval or i + interval + 1 > len(data):
            continue
        y.append(data[i])
        tp = []
        for j in range(13):
            if j == interval:
                continue
            tp.append(data[i + j - interval])
        x.append(tp)
    return (x, y)

x_train, y_train = generate_x_y(train)
x_test, y_test = generate_x_y(test)

#%%
from sklearn.linear_model import LinearRegression
clf = LinearRegression()
clf.fit(x_train, y_train)
y_pred = clf.predict(x_test)

diff = abs(y_pred - y_test)
print("avg: ", sum(diff) / len(diff))
print("max: ", max(diff))
print("min: ", min(diff))