Linux(Centos7.8)中conda虚拟环境搭建LSTM神经网络基于django3.1.2的api接口
2024-04-09 00:10:51
目录
- 1.准备工作
- 2.项目需求
- 2.1 根据需求下载LSTM依赖包
- 2.2 代码实现
- 3.启动服务
- 4、第三方应用调用webapi服务提供的api接口
1.准备工作
由上一博客Linux(Centos7.8)中Anaconda3使用虚拟环境搭建Django3应用Django应用的基础环境已搭建完备。
2.项目需求
第三方外部应用能够通过调用api接口的方式，获取由LSTM（长短期记忆）递归神经网络预测的未来30天的活期存款余额趋势数据。
2.1 根据需求下载LSTM依赖包
进入到由上篇博客搭建的虚拟环境目录/opt/anaconda3/envs/ml_api并激活环境
cd /opt/anaconda3/envs/ml_api
# 激活环境
conda activate ml_api
使用conda安装lstm神经网络依赖包,考虑到版本兼容性问题指定版本下载
conda install numpy
conda install matplotlib
conda install pandas
conda install pymysql
python -m pip install django-cors-headers #跨域
python -m pip install --upgrade --force-reinstall tensorflow==2.2.0 --ignore-installed
python -m pip install --upgrade --force-reinstall keras==2.3.0 --ignore-installed
python -m pip install --upgrade --force-reinstall sklearn --ignore-installed
conda list 查看该虚假环境已安装的包
(ml_api) [root@localhost api]# conda list
# packages in environment at /opt/anaconda3/envs/ml_api:
#
# Name Version Build Channel
_libgcc_mutex 0.1 main
_tflow_select 2.3.0 mkl
absl-py 0.11.0 py38h06a4308_0
aiohttp 3.6.3 py38h7b6447c_0
asgiref 3.2.10 pypi_0 pypi
astunparse 1.6.3 py_0
async-timeout 3.0.1 py38_0
attrs 20.2.0 py_0
blas 1.0 mkl
blinker 1.4 py38_0
brotlipy 0.7.0 py38h7b6447c_1000
c-ares 1.16.1 h7b6447c_0
ca-certificates 2020.10.14 0
cached-property 1.5.2 pypi_0 pypi
cachetools 4.1.1 py_0
certifi 2020.6.20 pyhd3eb1b0_3
cffi 1.14.3 py38he30daa8_0
chardet 3.0.4 py38_1003
click 7.1.2 py_0
cryptography 3.1.1 py38h1ba5d50_0
cycler 0.10.0 py38_0
dbus 1.13.18 hb2f20db_0
django 3.1.2 py_0
django-cors-headers 3.5.0 pypi_0 pypi
expat 2.2.10 he6710b0_2
fontconfig 2.13.0 h9420a91_0
freetype 2.10.4 h5ab3b9f_0
gast 0.3.3 py_0
glib 2.66.1 h92f7085_0
google-auth 1.23.0 pyhd3eb1b0_0
google-auth-oauthlib 0.4.2 pyhd3eb1b0_2
google-pasta 0.2.0 py_0
grpcio 1.31.0 py38hf8bcb03_0
gst-plugins-base 1.14.0 hbbd80ab_1
gstreamer 1.14.0 hb31296c_0
h5py 3.0.0 pypi_0 pypi
hdf5 1.10.6 hb1b8bf9_0
icu 58.2 he6710b0_3
idna 2.10 py_0
importlib-metadata 2.0.0 py_1
intel-openmp 2020.2 254
joblib 0.17.0 pypi_0 pypi
jpeg 9b h024ee3a_2
keras 2.3.0 pypi_0 pypi
keras-applications 1.0.8 pypi_0 pypi
keras-preprocessing 1.1.2 pypi_0 pypi
kiwisolver 1.3.0 py38h2531618_0
krb5 1.17.1 h173b8e3_0
lcms2 2.11 h396b838_0
ld_impl_linux-64 2.33.1 h53a641e_7
libedit 3.1.20191231 h14c3975_1
libffi 3.3 he6710b0_2
libgcc-ng 9.1.0 hdf63c60_0
libgfortran-ng 7.3.0 hdf63c60_0
libpng 1.6.37 hbc83047_0
libpq 12.2 h20c2e04_0
libprotobuf 3.13.0.1 hd408876_0
libstdcxx-ng 9.1.0 hdf63c60_0
libtiff 4.1.0 h2733197_1
libuuid 1.0.3 h1bed415_2
libxcb 1.14 h7b6447c_0
libxml2 2.9.10 hb55368b_3
lz4-c 1.9.2 heb0550a_3
markdown 3.3.2 py38_0
matplotlib 3.3.2 0
matplotlib-base 3.3.2 py38h817c723_0
mkl 2020.2 256
mkl-service 2.3.0 py38he904b0f_0
mkl_fft 1.2.0 py38h23d657b_0
mkl_random 1.1.1 py38h0573a6f_0
multidict 4.7.6 py38h7b6447c_1
ncurses 6.2 he6710b0_1
numpy 1.19.4 pypi_0 pypi
numpy-base 1.19.2 py38hfa32c7d_0
oauthlib 3.1.0 py_0
olefile 0.46 py_0
openssl 1.1.1h h7b6447c_0
opt_einsum 3.1.0 py_0
pandas 1.1.3 py38he6710b0_0
pcre 8.44 he6710b0_0
pillow 8.0.1 py38he98fc37_0
pip 20.2.4 py38_0
protobuf 3.13.0.1 py38he6710b0_1
psycopg2 2.8.5 py38h1ba5d50_0
pyasn1 0.4.8 py_0
pyasn1-modules 0.2.8 py_0
pycparser 2.20 py_2
pyjwt 1.7.1 py38_0
pymysql 0.10.1 py38_0
pyopenssl 19.1.0 py_1
pyparsing 2.4.7 py_0
pyqt 5.9.2 py38h05f1152_4
pysocks 1.7.1 py38_0
python 3.8.3 hcff3b4d_2
python-dateutil 2.8.1 py_0
pytz 2020.1 py_0
pyyaml 5.3.1 pypi_0 pypi
qt 5.9.7 h5867ecd_1
readline 8.0 h7b6447c_0
requests 2.24.0 py_0
requests-oauthlib 1.3.0 py_0
rsa 4.6 py_0
scikit-learn 0.23.2 pypi_0 pypi
scipy 1.5.3 pypi_0 pypi
setuptools 50.3.0 py38hb0f4dca_1
sip 4.19.13 py38he6710b0_0
six 1.15.0 py_0
sklearn 0.0 pypi_0 pypi
sqlite 3.33.0 h62c20be_0
sqlparse 0.4.1 py_0
tensorboard 2.3.0 pyh4dce500_0
tensorboard-plugin-wit 1.6.0 py_0
tensorflow 2.2.0 mkl_py38h6d3daf0_0
tensorflow-base 2.2.0 mkl_py38h5059a2d_0
tensorflow-estimator 2.2.0 pyh208ff02_0
termcolor 1.1.0 py38_1
threadpoolctl 2.1.0 pypi_0 pypi
tk 8.6.10 hbc83047_0
tornado 6.0.4 py38h7b6447c_1
urllib3 1.25.11 py_0
werkzeug 1.0.1 py_0
wheel 0.35.1 py_0
wrapt 1.12.1 py38h7b6447c_1
xz 5.2.5 h7b6447c_0
yarl 1.6.2 py38h7b6447c_0
zipp 3.4.0 pyhd3eb1b0_0
zlib 1.2.11 h7b6447c_3
zstd 1.4.5 h9ceee32_0
2.2 代码实现
settings.py
(ml_api) [root@localhost webapi]# vim settings.py ALLOWED_HOSTS = ['*']# Application definitionINSTALLED_APPS = ['django.contrib.admin','django.contrib.auth','django.contrib.contenttypes','django.contrib.sessions','django.contrib.messages','django.contrib.staticfiles','corsheaders', #跨域配置
]MIDDLEWARE = ['django.middleware.security.SecurityMiddleware','django.contrib.sessions.middleware.SessionMiddleware','django.middleware.common.CommonMiddleware','django.middleware.csrf.CsrfViewMiddleware','django.contrib.auth.middleware.AuthenticationMiddleware','django.contrib.messages.middleware.MessageMiddleware','django.middleware.clickjacking.XFrameOptionsMiddleware','corsheaders.middleware.CorsMiddleware',#跨域配置'django.middleware.common.CommonMiddleware',#跨域配置
]CORS_ORIGIN_ALLOW_ALL = True #跨域配置ROOT_URLCONF = 'webapi.urls'TEMPLATES = [{'BACKEND': 'django.template.backends.django.DjangoTemplates','DIRS': [os.path.join(BASE_DIR, "templates")],'APP_DIRS': True,'OPTIONS': {'context_processors': ['django.template.context_processors.debug','django.template.context_processors.request','django.contrib.auth.context_processors.auth','django.contrib.messages.context_processors.messages',],},},
]WSGI_APPLICATION = 'webapi.wsgi.application'# Database
# https://docs.djangoproject.com/en/3.1/ref/settings/#databasesDATABASES = {'default': {'ENGINE': 'django.db.backends.sqlite3','NAME': BASE_DIR / 'db.sqlite3',}
}# Password validation
# https://docs.djangoproject.com/en/3.1/ref/settings/#auth-password-validatorsAUTH_PASSWORD_VALIDATORS = [{'NAME': 'django.contrib.auth.password_validation.UserAttributeSimilarityValidator',},{'NAME': 'django.contrib.auth.password_validation.MinimumLengthValidator',},{'NAME': 'django.contrib.auth.password_validation.CommonPasswordValidator',},{'NAME': 'django.contrib.auth.password_validation.NumericPasswordValidator',},
]# Internationalization
# https://docs.djangoproject.com/en/3.1/topics/i18n/LANGUAGE_CODE = 'en-us'TIME_ZONE = 'UTC'USE_I18N = TrueUSE_L10N = TrueUSE_TZ = True# Static files (CSS, JavaScript, Images)
# https://docs.djangoproject.com/en/3.1/howto/static-files/STATIC_URL = '/static/'在项目下新建视图文件 views.py
touch view.py
# -*- coding: utf-8 -*-
import json
from django.http import HttpResponse
from django.views.decorators.csrf import csrf_exempt
import numpy as np
import matplotlib.pyplot as plt
import pandas as pd
from sklearn.preprocessing import MinMaxScaler
from tensorflow.keras.models import Sequential # Sequential 用于初始化神经网络
from tensorflow.keras.layers import Dense # Dense 用于添加全连接的神经网络层
from tensorflow.keras.layers import LSTM # LSTM 用于添加长短期内存层
from tensorflow.keras.layers import Dropout # Dropout 用于添加防止过拟合的dropout层
import os
import datetime
import matplotlib.dates as mdates
import matplotlib.pyplot as plt
import tensorflow as tf
#os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'
import pymysql.cursors# 连接数据库
connect = pymysql.Connect(host='192.168.6.100',port=3306,user='root',passwd='123456',db='test',charset='utf8'
)def plot_show(axis_X, real_alive_amt, predicted_stock_price):""":param axis_X::param real_alive_amt::param predicted_stock_price::return: z展示图"""plt.rcParams['font.sans-serif'] = ['SimHei']plt.rcParams['axes.unicode_minus'] = Falseplt.plot(axis_X, real_alive_amt, color='green', label='真实活期余额')# 显示活期余额plt.plot(axis_X, predicted_stock_price[:, -1], ls='--', color='red', label='预测活期余额')plt.title('活期余额预测')plt.gca().xaxis.set_major_locator(mdates.DayLocator())plt.xlabel('Time')plt.ylabel('余额（元）')plt.legend()plt.tick_params(axis='both', which='major', labelsize=16)plt.gcf().autofmt_xdate() # 自动旋转日期标记plt.show()def train_data(dataset_train,step_len):"""构建训练数据:param training_set::return: X_train, y_train, sc"""sc = MinMaxScaler(feature_range=(0, 1))training_set = dataset_train.valuestraining_set_scaled = sc.fit_transform(training_set)X_train = []y_train = []for i in range(step_len, len(training_set_scaled)):X_train.append(training_set_scaled[i - step_len:i, :])y_train.append(training_set_scaled[i, :])X_train, y_train = np.array(X_train), np.array(y_train)X_train = np.reshape(X_train, (X_train.shape[0], X_train.shape[1], X_train.shape[2]))return X_train, y_train, scdef test_data(dataset_train, dataset_test, sc,step_len):"""构建测试数据:param dataset_train::param dataset_test::param sc::return: X_test"""# 合并训练集和测试集dataset_total = pd.concat((dataset_train, dataset_test), axis=0)db_all = dataset_total[:].valuesnp.random.shuffle(db_all) # shuffle() 方法将序列的所有元素随机排序。inputs = db_all[len(dataset_total) - len(dataset_test) - int(step_len):]inputs = inputs.reshape(-1, dataset_test.shape[1])inputs = sc.transform(inputs)X_test = []for i in range(int(step_len), int(step_len) + len(dataset_test)):curr_frame = inputs[i - int(step_len):i - 1, :]curr_frame = np.insert(curr_frame, step_len-1,dataset_test.values, axis=0)X_test.append(curr_frame)X_test = np.array(X_test)X_test = np.reshape(X_test, (X_test.shape[0], X_test.shape[1], X_test.shape[2]))return X_testdef bulid_model(X_train, y_train):"""创建股票预测模型:param X_train::param y_train::return: regressor"""regressor = Sequential()# LSTM的输入为 [samples, timesteps, features],这里的timesteps为步数，features为维度 这里我们的数据是11维的regressor.add(LSTM(units=50, return_sequences=True, input_shape=(X_train.shape[1], 11)))regressor.add(Dropout(0.2))regressor.add(LSTM(units=50, return_sequences=True))regressor.add(Dropout(0.2))regressor.add(LSTM(units=50, return_sequences=True))regressor.add(Dropout(0.2))regressor.add(LSTM(units=50))regressor.add(Dropout(0.2))# 全连接，输出11个regressor.add(Dense(units=11))regressor.compile(optimizer='adam', loss='mean_squared_error')regressor.fit(X_train, y_train, epochs=1, batch_size=60)#print(regressor.summary())return regressordef predict_test_data_fun(dataset_train,regressor,sc,X_test,real_alive_amt,step_len,predict_custs):"""最新预测数据:param dataset_train::param regressor::param sc::param X_test::param real_alive_amt::param step_len::return:"""predicted_alive_amt = regressor.predict(X_test)# 反向转换trans_predicted_alive_amt = sc.inverse_transform(predicted_alive_amt)predict_custs.insert(step_len,trans_predicted_alive_amt[-1])# 添加表头result_dataframe = pd.DataFrame(predicted_alive_amt, columns=['idcard_first_6','idcard_middle_8','idcard_last_4','gender','age','alive_avg_amt','c_alive_amt', 'd_alive_amt', 'is_arrear','arrear_amt','total_alive_amt'])return test_data(dataset_train,result_dataframe,sc,step_len),trans_predicted_alive_amt[:, -1],predict_custsdef recursion_call(dataset_train,regressor,sc,X_test,real_alive_amt,step_len,recursion_day,predict_custs):"""预测指定天数函数:param dataset_train::param regressor::param sc::param X_test::param real_alive_amt::param step_len::param recursion_day::return:"""if recursion_day == 0:return 0else:X_test,real_alive_amt,predict_custs = predict_test_data_fun(dataset_train,regressor,sc,X_test,real_alive_amt,step_len,predict_custs)return recursion_call(dataset_train,regressor,sc,X_test,real_alive_amt,step_len,recursion_day - 1,predict_custs)#@csrf_exempt
def alive_api(request,custno):print(custno)custno=int(custno)result = {"result": 0, "msg": "执行成功"}# 读取文件df = pd.read_csv('/opt/anaconda3/envs/ml_api/api/webapi/alives.csv',parse_dates=['dt'])df['idcard_last_4'] =df['idcard_last_4'].apply(lambda x:str(x).replace('X','10').zfill(4))# 根据客户号去重cust_ids = df.drop_duplicates(subset=['customer_no'], keep='first')['customer_no']# 步长step_len = 60# 递归天数recursion_day = 30#custno=1000340109# 训练集dataset_train = df[(df['dt'] < '2020-10-18') & (df['customer_no'] == custno)].loc[:,['idcard_first_6','idcard_middle_8','idcard_last_4','gender','age','alive_avg_amt', 'c_alive_amt','d_alive_amt','is_arrear','arrear_amt','total_alive_amt']]# 测试集dataset_test = df[(df['dt'] == '2020-10-18') & (df['customer_no'] == custno)].loc[:,['idcard_first_6','idcard_middle_8','idcard_last_4','gender','age','alive_avg_amt', 'c_alive_amt','d_alive_amt','is_arrear','arrear_amt','total_alive_amt']]# 获取存款余额真实值real_alive_amt = dataset_test['total_alive_amt'].values# 训练数据X_train, y_train, sc = train_data(dataset_train,step_len)# 构建模型regressor = bulid_model(X_train, y_train)# 构造测试数据X_test = test_data(dataset_train, dataset_test, sc,step_len)# 插入数据predict_custs = list()recursion_call(dataset_train,regressor,sc,X_test,real_alive_amt,step_len,recursion_day,predict_custs)#print(predict_custs)jsonObj ={}for index in range(len(predict_custs)):key = ((datetime.datetime.now() + datetime.timedelta(days=index)).strftime("%Y-%m-%d"))jsonObj[str(key)]=str(predict_custs[index][-1])#json返回为中文#print(jsonObj)return HttpResponse(json.dumps(jsonObj,ensure_ascii=False),content_type="application/json,charset=utf-8")urls.py
from django.contrib import admin
from django.urls import path
from django.conf.urls import url
from . import viewsurlpatterns = [path('admin/', admin.site.urls),url(r'^api/(?P<custno>\d+)/$',views.alive_api,name='alive_api'),
]
核心代码到此为止。
3.启动服务
python manage.py runserver 0.0.0.0:8000
4、第三方应用调用webapi服务提供的api接口
图标展示采用echart.js 核心代码如下：
服务接口：http://192.168.2.9:8000/api/1001324295/，其中1001324295为参数
<div id="main" style="width: 1200px;height:400px;"></div>
<script type="text/javascript">$(function(){xdata=[]ydata=[]var myChart = echarts.init(document.getElementById('main'),'macarons');myChart.showLoading({text: '大数据正在预测...',textStyle: { fontSize : 30 , color: '#444' },effectOption: {backgroundColor: 'rgba(0, 0, 0, 0)'}});$.get('http://192.168.2.9:8000/api/1001324295/',function(dic) {console.log(dic)$.each(dic,function(index,item){console.log(index+"|"+item)xdata.push(index)ydata.push(item)})//console.log(xdata)//console.log(ydata)// 基于准备好的dom，初始化echarts实例// 指定图表的配置项和数据var option = {title: {text: '未来30天活期存款余额预测趋势图',textStyle:{color:'#0DB9F2', //颜色fontStyle:'normal', //风格fontWeight:'normal', //粗细fontFamily:'Microsoft yahei', //字体fontSize:24, //大小align:'center' //水平对齐},x:'center'},tooltip: {},legend: {data:['未来30天活期存款余额预测趋势图']},xAxis: {data: xdata,axisLabel: {interval:0,rotate:50}},yAxis: {type: 'value',//默认以千分位显示，不想用的可以在这加一段axisLabel : { //调整左侧Y轴刻度， 直接按对应数据显示show:true,showMinLabel:true,showMaxLabel:true,formatter: function (value) {return value;}}},series: [{name: '活期存款余额（元）',type: 'line',data: ydata,markPoint: {itemStyle: {normal: {borderColor: '#87cefa',borderWidth: 1, // 标注边线线宽，单位px，默认为1label: {show: true}}},data: [{type: 'max',name: '最大值'},{type: 'min',name: '最小值'}]},markLine: {data: [{type: 'average', name: '平均值'},[{symbol: 'none',x: '90%',yAxis: 'max'}, {symbol: 'circle',label: {normal: {position: 'start',formatter: '最大值'}},type: 'max',name: '最高点'}]]}}]};// 使用刚指定的配置项和数据显示图表。myChart.setOption(option);myChart.hideLoading();});});</script>第三方应用调用预测api接口，由预测模型需要递归多次处理时间比较久，加载过程中提示loading为大数据正在预测…
Linux(Centos7.8)中conda虚拟环境搭建LSTM神经网络基于django3.1.2的api接口相关推荐
- 忆阻器交叉开关阵列中的长短期记忆(LSTM)神经网络
忆阻器交叉开关阵列中的长短期记忆(LSTM)神经网络 原文:Long short-term memory networks in memristor crossbar arrays 作者:CanLi. ...
- 关于vue2用vue-cli搭建环境后域名代理的http-proxy-middleware解决api接口跨域问题
在vue中用http-proxy-middleware来进行接口代理,比如:本地运行环境为http://localhost:8080但真实访问的api为 http://www.baidu.com这时我 ...
- 腾讯云服务器linux+CentOS7.9+yum源+nginx搭建网站
第一次建网站 文章目录 前言 买服务器 配置服务器 使用xshell 连接 写一个html网页 创建文件夹 访问网站 安装nginx yum源配置 安装wegt 安装nginx 前言 参考B站程序员鱼 ...
- linux centos7.9中安装docker的3种方式-图解
使用华为云或腾讯云,阿里云的linux centos 7.9版本,都是自带yum 命令的. 最快的方式shell脚本安装 shell脚本方式安装 curl -fsSL https://get.dock ...
- linux配置中心自动化,centos下搭建git服务 基于git服务配置自动化部署 遇到的Linux知识补充...
安装 git su root yum install -y git 添加 git 用户,用于管理/使用 git 服务/仓库 创建用户 git # 添加名为git的用户, 并且会默认创建一个名为git的 ...
- 搭建股票数据api接口过程中会遇到什么问题？
搭建股票数据api接口过程中会遇到什么问题?以下这些问题比较常见,对于可能出现问题,建议每个搭建者最好能够提前了解并掌握更好的解决方案,从而应对可能出现的潜在问题. 拉合作方还是推数据? 股票数据ap ...
- API接口自动化测试框架搭建（一）-总体需求
(一)-总体需求 1 实现目的 2 功能需求 3 其他要求 4 适用人员 5 学习周期 6 学习建议 7 内容直达 8 反馈联系 1 实现目的 API接口自动化测试,主要针对http接口协议: 便于回 ...
- php360随机壁纸图片接口,PHP超简单的随机图片头像API接口网站搭建教程
摘要 向上文所说,为了其他邮箱调用随机图片,我当时特意去网上找了些随机图片API接口网站,发现他们API接口有点慢,本着不外链的精神,自己搭建了个随机图片and头像的API接口.在这里把教程分享给大家 ...
- Linux CentOS7 中 完美解决VMTools失效，windows 与 Liunx间完美复制文件，无报错的解决方案
Linux CentOS7 中 完美解决VMTools失效,windows 与 Liunx间完美复制文件,无报错的解决方案 DreamFollowerss 2020-05-25 12:03:08 3 ...
最新文章
- 设置用户帐号只能一个地方登录
- 带着问题读CLR via C#（六）常量与字段
- Winform中设置ZedGraph多条Y轴时坐标轴左右显示设置
- android bitmap转图片_Android 这些 Drawable 你都会用吗？
- 网络协议之:一定要大写的SOCKS
- 京东物流研发岗位会背景调查吗_【秋招资讯】京东健康于港交所主板上市 | 京东健康2021校园招聘火热进行中！...
- [软件]提高日常操作电脑的效率
- Undefined control sequence.l.113 \LinesNumbered
- perl hash输出类似HASH(0x26caf78)
- web前端简历个人技能该怎么写?
- kindle mysql_MySQL(二) MySQL基本操作
- 远程控制工具ToDesk测评
- c语言 数组 全部赋值0,关于C语言数组赋值方法
- 电脑cpu电压测试软件,CPU-Z(cpu检测工具)
- 不同公式等号对齐_魔方的公式（第二版）
- 硅芯片拆解：早期 555 定时器芯片内部一览
- Makefile与前缀后缀相关函数suffix、basename、addsuffix、addprefix
- Solved - Safari keeps forcing HTTPS on localhost
- Mac删除文件提示“不能删除xx项目，正在使用中”
- PROE5.0飞利浦耳机结构设计教程