kaggle数据分析家庭电力消耗过程详解

!pip install prophet  -i https://pypi.tuna.tsinghua.edu.cn/simple

data_path="/home/mw/input/Household_Electricity4767/household_power_consumption.csv"

import pandas as pd
import seaborn as sns
import numpy as np
from tqdm.auto import tqdm
from prophet import Prophet

df=pd.read_csv(data_path)

df.head()

df.describe()

df.dtypes

index                      int64
Date                      object
Time                      object
Global_active_power       object
Global_reactive_power     object
Voltage                   object
Global_intensity          object
Sub_metering_1            object
Sub_metering_2            object
Sub_metering_3           float64
dtype: object

df['Date']=pd.DatetimeIndex(df['Date'])

make_em_num = ['Global_active_power', 'Global_reactive_power', 'Voltage', 'Global_intensity', 'Sub_metering_1', 'Sub_metering_2', 'Sub_metering_3']

def floating(string):
    try:
        return float(string)
    except:
        return float(0)

for column in tqdm(make_em_num):
    df[column] = df[column].apply(lambda item: floating(item))

HBox(children=(FloatProgress(value=0.0, max=7.0), HTML(value='')))

df.dtypes

index                             int64
Date                     datetime64[ns]
Time                             object
Global_active_power             float64
Global_reactive_power           float64
Voltage                         float64
Global_intensity                float64
Sub_metering_1                  float64
Sub_metering_2                  float64
Sub_metering_3                  float64
dtype: object

df.head()

sns.heatmap(df.drop(['index','Date','Time'], axis=1).corr(), annot=True)

<matplotlib.axes._subplots.AxesSubplot at 0x7f31603ed4e0>

import matplotlib.pyplot as plt

df.shape

(260640, 10)

df=df.sample(n=10000)

def prophet_forecaster(data, x, y, period=100):
    new_df = pd.DataFrame(columns=['ds', 'y'])
    new_df['ds']= data[x]
    new_df['y'] = data[y]
    model = Prophet()
    model.fit(new_df)
    future_dates = model.make_future_dataframe(periods=period)
    forecast = model.predict(future_dates)
    model.plot(forecast)
    plt.title(f"Forecasting on the next {period} days for {y}")

prophet_forecaster(df, x='Date', y='Global_active_power', period=100)

prophet_forecaster(df, x='Date', y='Voltage', period=100)

INFO:prophet:Disabling yearly seasonality. Run prophet with yearly_seasonality=True to override this.
INFO:prophet:Disabling daily seasonality. Run prophet with daily_seasonality=True to override this.