qiulianglee / informer_chinesenotes

informer_ChineseNotes

informer_ChineseNotes has added Chinese annotations and some visual additions to the informer2020 project.

使用场景

• 时间序列预测
• 股票预测
• 市场预测
• 电力预测
• 天气预测
• ....

Requirements

• torch_lts
• pyecharts
• Python 3.7
• matplotlib
• numpy == 1.19.4
• pandas
• scikit_learn
• xlrd
• ....

You can install it directly from the environment pack version file:

pip install -r requirements.txt
or
conda env create -f torch_lts.yaml

Quick Configuration Environment

Method 1：

Create a virtual environment in Anaconda Powershell Prompt (Anaconda3) :

conda create -n torch_lts python=3.7

Then enter the virtual environment:

activate torch_lts

Finally, install the environment dependency package using the requirements. TXT file:

pip install -r requirements.txt

Method 2

Create virtual environments and install dependencies using conda and YAMl files directly from Anaconda Powershell Prompt (Anaconda3) :

conda env create -f torch_lts.yaml

Method 3

Execute the following commands directly in your environment to install:

conda install pytorch torchvision torchaudio cudatoolkit=11.1 -c pytorch-lts -c conda-forge

pip install numpy == 1.19.4
pip install pyecharts
pip install xlrd
pip install openpyxl
pip install matplotlib
pip install pandas

Import virtual environment in PyCharm:

When selecting the interpreter, do not create a new virtual environment and use the virtual environment you already created:

FAQ

If you run into a problem like RuntimeError: The size of tensor a (98) must match the size of tensor b (96) at non-singleton dimension 1, you can check torch version or modify code about Conv1d of TokenEmbedding in models/embed.py as the way of circular padding mode in Conv1d changed in different torch versions.

Interesting and useful tool recommendations

• streamlit

Thank you

• Thanks to @zhouhaoyi open source algorithm paper and code repository.

reference

• Original repository: informer2020
• Zhouhaoyi paper: Informer：Beyond Efficient Transformer for Long Sequence Time-Series Forecasting (AAAI'21 Best Paper)

Tips

数据是网上找来的，已做脱敏处理，如有冒犯到您的请联系e-mail或者从Tony's blog中私聊我侵删。

To contact me

• e-mail：[email protected]