[nilm] non intrusive load decomposition module nilmtk installation tutorial

         This tutorial focuses on Anaconda Of Python Installation in the environment nilmtk modular , Therefore, it is recommended to install nilmtk Install the module before Anaconda Software , The download link can be found in Download link . install nilmtk Modules are recommended in anconda Create in a virtual environment , So we need to use Conda Command to create a virtual environment , Download again nitmtk modular , install nilmtk modular .

1 conda Command to create a virtual environment

1.1 conda Tutorial on the use of commands

• Get version number

conda --version/conda -V

• get help

conda --help
conda -h
# View help for a command , Such as update Orders and remove command 
conda update --help
conda remove --help
# Empathy , Of the above commands --help It can also be replaced by -h.

• Environmental management

# View all command help for Environment Management 
conda env -h
# establish python A virtual environment 
conda create --name your_env_name #( Here you can rewrite the name of your virtual environment )

Input y Confirm creation .

• Create formulation python Version environment

# Create formulation python Version environment 
conda create --name your_env_name python=2.7
conda create --name your_env_name python=3
conda create --name your_env_name python=3.5
# Create environments that contain certain packages 
conda create --name your_env_name numpy scipy
#  List all current environments 
conda info --envs
conda env list
#  Enter an environment 
conda activate your_env_name# This is recommended 
activate your_env_name
# Exit the current environment 
conda deactivate
# Copy an environment 
conda create --name new_env_name --clone old_env_name 
# Delete an environment 
conda remove --name your_env_name --all
# From yml File creation environment 
conda env create -f environment.yml
# List all packages in the currently active environment 
conda list
# List all packages in a given environment 
conda list -n your_env_name

2 install nilmtk modular

2.1 download nilmtk modular

from github Download directly from nilmtk Module installation package , Download link https://github.com/nilmtk/nilmtk . Unzip the file

2.2 open Anaconda Prompt

Switch the current path to nilmtk In the decompression path , The switching command adopts cd Command switching , Before switching paths, you need to switch to the disk where the pressurized folder is located , Then switch to the specific path (C If you use a disc, you can use it directly cd Switch ）. for example ‘E:\ experiment \ modular \nilmtk’ In the path , I'm going to switch to E disc , Switch to ‘E:\ experiment \ modular \nilmtk’ in . The order is as follows ：

# Switch to E disc 
E:
# Switch to specific path 
cd E:\ experiment \ modular \nilmtk

2.3 structure Nilmtk Virtual environment for

This step mainly builds nilmtk Virtual environment for nilmtk-env, If it has been established in other ways nilmtk-env, You need to delete the directory first , Otherwise, the operation fails , The first part of the deletion method has already introduced . The create command is as follows ：

conda env create -f environment.yml

2.4 Switch to nilmtk Virtual environment for nilmtk-env

Build complete specifically for nilm Module installation python Environmental Science , Need to switch to nilm-env Environment , To use the installed nilmtk modular ,anaconda The modules of the module must be in nilmtk-env The environment exists , Need to switch to nilmtk-env in , use pip install  XXX Command to install . Specific commands

conda activate nilmtk-env
# perhaps 
activate nilmtk-env
# Out of the environment 
conda deactivate

 2.5 install nilmtk modular

Switch to nillmtk-env After the virtual environment, enter the corresponding command to install nilmtk modular , In this way, you can also nilmk-metadata Install well .

python setup.py develop

After running the command , Will run for a while , Wait for the run to finish , If no error is reported during the operation, it will be ok,

After running, use the following two commands to test .

conda list
nosetests

  If conda list The above modules are displayed , Indicates successful installation .

You can test it with the following code ,low_freq Express redd The original path of the dataset , among REDD Dataset address ：http://redd.csail.mit.edu/, user name ： redd  ; password ： disaggregatetheenergy

from nilmtk.dataset_converters import convert_redd
convert_redd(r'D:\MyProject\NILMTK\low_freq',r'D:\MyProject\NILMTK\low_freq\redd.h5')

If the operation is successful, the installation is successful .

3 install Spyder Editor

anaconda Bring one with you. spyder Editor , To write python The program is very convenient , however nilmtk-env The module does not spyder Tools , In order to write nilmtk Related codes , Run the load decomposition program , Need to be in nilmtk-env Environmental installation spyder. Use command ：

conda install spyder

Can pass conda list Or view the folder directory in the start menu bar . The effect is shown below .

 4 install nilmtk-contrib modular

This module contains the use of NILMTK A quick experiment of API All state-of-the-art algorithms for the energy decomposition task implemented . The installation method is the same as nilmtk The modules are the same , To download nilmtk-contrib Module installation package , Then unzip the installation package , stay nilmtk-env Switch to nilmtk-contrib Folder , function python setup.py develop, And then run conda list Check whether the installation is successful .

nilmtk-contrib Installation package download path ：https://github.com/nilmtk/nilmtk-contrib

5 Installation package download

If github Unable to access download , You can go to the following link to download nilmtk and nilmtk-contrib Etc .

https://download.csdn.net/download/weixin_56516468/85698419

【 Reference material 】

1.https://blog.csdn.net/wangwy_hoveringeagle/article/details/121177790

2.https://blog.csdn.net/wangwy_hoveringeagle/article/details/121143285

3.https://blog.csdn.net/weixin_56516468/article/details/118058994?utm_medium=distribute.pc_relevant.none-task-blog-2~default~baidujs_baidulandingword~default-1-118058994-blog-121143285.pc_relevant_paycolumn_v3&spm=1001.2101.3001.4242.2&utm_relevant_index=4