利用Prophet来预测气温

本文介绍了如何运用Facebook的Prophet库对气温数据进行时间序列预测。通过详细步骤,展示了如何加载数据、预处理、建立模型以及解读预测结果,帮助读者理解并掌握Prophet在气象预测中的应用。
摘要由CSDN通过智能技术生成

import pandas as pd
import numpy as np
from fbprophet import Prophet

df = pd.read_csv('tq.txt',header=None,sep='\\s+',usecols=[0,2,])
#print(df.shape)
#print(df.head())

print(type(df))

#print(df[0])
df['ds']=df[0]
df['y'] =df[2]

m = Prophet()
m.fit(df);

future = m.make_future_dataframe(periods=30)
#print(future.tail())
forecast = m.predict(future)
print(forecast)






数据如下

2015-07-17  26  34
2015-07-18  26  34
2015-07-19  25  34
2015-07-21  25  28 
2015-07-22  25  29 
2015-07-24  26  33 
2015-07-25  26  33
2015-07-26  27  36 
2015-07-27  27  36 
2015-07-28  26  36 
2015-07-29  25  34 
2015-07-30  25  35 
2015-07-31  25  36 
2015-08-01  26  36 
2015-08-02  26  36 
2015-08-03  26  36 
2015-08-04  26  36 
2015-08-05  25  35 
2015-08-06  26  35 
2015-08-07  26  32 
2015-08-10  25  30 
2015-08-11  26  33 
2015-08-12  26  34 
2015-08-13  26  34 
2015-08-14  26  33 
2015-08-15  26  33 
2015-08-16  26  33 
2015-08-18  26  32 
2015-08-20  25  34 
2015-08-21  25  33 
2015-08-22  26  32 
2015-08-23  26  32 
2015-08-24  25  33 
2015-08-25  25  34 
2015-08-26  25  33 
2015-08-27  25  30 
2015-08-28  25  30 
2015-08-29  24  28 
2015-08-30  23  27 
2015-08-31  23  29 
2015-09-01  23  27 
2015-09-02  24  28 
2015-09-03  24  32 
2015-09-04  25  32 
2015-09-05  25  34 
2015-09-06  24  33 
2015-09-07  23  31 
2015-09-08  23  30 
2015-09-09  23  29 
2015-09-10  22  29 
2015-09-11  22  31 
2015-09-12  23  30 
2015-09-13  23  26 
2015-09-14  23  26 
2015-09-15  23  28 
2015-09-16  23  28 
2015-09-17  23  29 
2015-09-18  23  29 
2015-09-19  23  27 
2015-09-20  20  24 
2015-09-22  22  26 
2015-09-23  24  32 
2015-09-24  23  29 
2015-09-25  23  28 
2015-09-26  22  26 
2015-09-27  23  30 
2015-09-28  23  29 
2015-09-29  24  27 
2015-09-30  24  30 
2015-10-01  22  31 
2015-10-02  21  25 
2015-10-03  22  27 
2015-10-04  24  28 
2015-10-05  24  27 
2015-10-06  23  27 
2015-10-07  22  26 
2015-10-08  21  27 
2015-10-09  21  23 
2015-10-10  20  24 
2015-10-11  18  25 
2015-10-12  17  26 
2015-10-13  17  25 
2015-10-14  17  25 
2015-10-15  18  27 
2015-10-16  18  27 
2015-10-17  18  27 
2015-10-18  20  26 
2015-10-19  21  25 
2015-10-20  21  24 
2015-10-21  21  26 
2015-10-22  20  28 
2015-10-24  19  28 
2015-10-25  19  26 
2015-10-26  19  28 
2015-10-27  20  29 
2015-10-30  19  24 
2015-10-31  16  19 
2015-11-01  16  19 
2015-11-03  17  20 
2015-11-04  20  23 
2015-11-05  21  25 
2015-11-06  22  28 
2015-11-07  22  31 
2015-11-08  22  30 
2015-11-09  18  23 
2015-11-10  17  20 
2015-11-11  17  20 
2015-11-12  17  20 
2015-11-13  19  26 
2015-11-14  18  22 
2015-11-15  19  24 
2015-11-16  20  25 
2015-11-17  21  28 
2015-11-18  20  29 
2015-11-19  19  24 
2015-11-20  20  25 
2015-11-21  20  24 
2015-11-22  19  24 
2015-11-23  18  26 
2015-11-24  17  25 
2015-11-25  12  21 
2015-11-26  9  17 
2015-11-27  10  16 
2015-11-28  12  18 
2015-11-29  15  19 
2015-11-30  16  18 
2015-12-01  16  20 
2015-12-02  15  22 
2015-12-03  13  18 
2015-12-04  13  17 
2015-12-05  11  15 
2015-12-06  10  15 
2015-12-07  12  17 
2015-12-08  13  16 
2015-12-09  14  16 
2015-12-10  13  21 
2015-12-11  13  18 
2015-12-12  13  17 
2015-12-13  13  17 
2015-12-14  14  17 
2015-12-15  10  16 
2015-12-16  7  15 
2015-12-17  5  12 
2015-12-18  6  13 
2015-12-19  10  15 
2015-12-20  13  18 
2015-12-21  15  18 
2015-12-22  15  18 
2015-12-23  15  18 
2015-12-24  13  17 
2015-12-25  12  15 
2015-12-26  12  15 
2015-12-27  10  14 
2015-12-28  9  15 
2015-12-29  9  14 
2015-12-30  10  15 
2015-12-31  9  15 
2016-01-01  10  16 
2016-01-02  12  21 
2016-01-03  14  21 
2016-01-04  15  22 
2016
评论 1
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值