产品定价 立即试用
Trendz Analytics
社区版 专业版 云平台 边缘 PE边缘 IoT网关 许可证服务器 Trendz Analytics 移动应用 PE移动应用 MQTT Broker PE MQTT Broker
文档 > 预测 > 自定义Python模型
入门
指南 安装 如何激活Trendz
目录

使用 Python 自定义模型

可通过编写自定义 Python 代码向 Trendz 添加新的预测模型。该代码在服务端执行，可访问完整输入数据集，包括所需的遥测与属性数据。 可在代码中导入所需的 Python 库，并基于输入数据预测目标指标。

多变量 Python 模型示例

本模板演示如何在 Trendz 中使用 Python 创建并实现自定义多变量预测模型。自定义模型可扩展平台内置预测能力，支持特定算法、附加变量或参数调优，以满足独特业务需求。

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103

#####################################################
# 预测方法：线性回归

from sklearn.preprocessing import MinMaxScaler, StandardScaler
from sklearn.linear_model import LinearRegression
import pickle
import numpy as np
import os 

class CustomModel(IModel):

    def __init__(self, value_transformer=None, timestamp_transformer=None):
        self.model = None
        self.timestamp_transformer = timestamp_transformer if timestamp_transformer else StandardScaler()
        self.value_transformer = value_transformer if value_transformer else MinMaxScaler()
        self.sum_x = 0
        self.sum_y = 0
        self.sum_xy = 0
        self.sum_xx = 0
        self.n = 0

    def init_state(self):
        self.model = LinearRegression()

    def train(self, data, additionalData=None):
        # Prepare
        ts = np.array([point[0] for point in data]).reshape(-1, 1)
        values = np.array([point[1] for point in data]).reshape(-1, 1)
        self.timestamp_transformer.fit(ts)
        self.value_transformer.fit(values)
        ts_scaled = self.timestamp_transformer.transform(ts)
        values_scaled = self.value_transformer.transform(values)

        # Fit
        self.sum_x = np.sum(ts_scaled)
        self.sum_y = np.sum(values_scaled)
        self.sum_xy = np.sum(ts_scaled * values_scaled)
        self.sum_xx = np.sum(ts_scaled ** 2)
        self.n = len(ts_scaled)

        self.model.fit(ts_scaled, values_scaled)

    def partial_fit(self, data, additionalData=None):
        # Prepare
        ts = np.array([point[0] for point in data]).reshape(-1, 1)
        values = np.array([point[1] for point in data]).reshape(-1, 1)
        # self.timestamp_transformer.partial_fit(ts)
        # self.value_transformer.partial_fit(values)
        ts_scaled = self.timestamp_transformer.transform(ts)
        values_scaled = self.value_transformer.transform(values)

        # Fit
        self.sum_x += np.sum(ts_scaled)
        self.sum_y += np.sum(values_scaled)
        self.sum_xy += np.sum(ts_scaled * values_scaled)
        self.sum_xx += np.sum(ts_scaled ** 2)
        self.n += len(ts_scaled)

        if self.n > 0:
            mean_x = self.sum_x / self.n
            mean_y = self.sum_y / self.n
            slope = (self.sum_xy - self.n * mean_x * mean_y) / (self.sum_xx - self.n * mean_x ** 2)
            intercept = mean_y - slope * mean_x
            self.model.coef_ = np.array([[slope]])
            self.model.intercept_ = np.array([intercept])

    def predict(self, timestamps):
        ts = np.array(timestamps).reshape(-1, 1)
        ts_scaled = self.timestamp_transformer.transform(ts)
        predictions_scaled = self.model.predict(ts_scaled)
        predictions = self.value_transformer.inverse_transform(predictions_scaled)
        return list(zip(timestamps, predictions.flatten()))

    def save_state(self, file_path):
        with open(file_path, 'wb') as file:
            state = {
                'model': self.model,
                'value_transformer': self.value_transformer,
                'timestamp_transformer': self.timestamp_transformer,
                'sum_x': self.sum_x,
                'sum_y': self.sum_y,
                'sum_xy': self.sum_xy,
                'sum_xx': self.sum_xx,
                'n': self.n
            }
            pickle.dump(state, file)

    def load_state(self, file_path):
        with open(file_path, 'rb') as file:
            state = pickle.load(file)
            self.model = state['model']
            self.value_transformer = state['value_transformer']
            self.timestamp_transformer = state['timestamp_transformer']
            self.sum_x = state['sum_x']
            self.sum_y = state['sum_y']
            self.sum_xy = state['sum_xy']
            self.sum_xx = state['sum_xx']
            self.n = state['n']

    def name(self):
        return "LinearRegressionModel"

#####################################################

下一步