The purpose of this project is for ease of setup VNPY trading application. 

With this project, the VNPY is running in a docker container, then user can use SSH to access the VNPY in the container. 

```shell
+--------------+                             +------------------------------------+
| Host machine |---> SSH (port 23456) --->   | Container (Ubuntu 18.04 with VNPY) |
+--------------+                             +------------------------------------+
```

It should work on other platforms, but currently it has been only tested on:

• Macbook Pro (Intel) with Docker Desktop.

For compiling successfully the ta-lib, it have to configure the 4GB+ memory resources.

* Pull from Docker Hub (recommended)

```shell
docker pull edwardyang/vnpy
```

```shell
$ docker images

REPOSITORY        TAG       IMAGE ID       CREATED          SIZE
edwardyang/vnpy   latest    7813668517cc   41 minutes ago   5.53GB
```

* Build from Dockerfile (it take hours)

Clone source code:

```shell
git clone https://github.com/Ed-Yang/vnpy-docker.git
```

Build VNPY container:

<yourname>: the user name (also used as default password) of Linux's account.

```shell
cd vnpy-docker
docker build -t vnpy --build-arg USERNAME=<yourname> .
```

```shell
$ docker images

docker images
REPOSITORY   TAG       IMAGE ID       CREATED         SIZE
vnpy         latest    7813668517cc   4 minutes ago   5.53GB
```

• Setup environment variable

  <root-of-repo>: the root path of this repo.
  <yourname>: the login name given in container building. (default: edward)

  cd vnpy-docker
  export VNPY_ROOT=`pwd`
  export USERNAME=<yourname>

• Setup working folder

  In local host:

  cp -r $VNPY_ROOT/examples $VNPY_ROOT/run

• Start VNPY container

  Pull from Docker Hub:

  docker run --rm --name vnpy -v $VNPY_ROOT/run:/home/$USERNAME/workspace/run -p 23456:23456 -it edwardyang/vnpy

  Build from Dockerfile:

  docker run --rm --name vnpy -v $VNPY_ROOT/run:/home/$USERNAME/workspace/run -p 23456:23456 -it vnpy

Currently, the VPNY container only support SSH connection.

• macOS

  First, install X windows client [XQuartz][https://www.xquartz.org].

  If the image is pulled from Docker Hub, the default username and password are 'edward'. Add an config "vnpy" in ~/.ssh/config:

  Host vnpy
  User <yourname>
  HostName localhost
  Port 23456
  SendEnv LANG LC_*
  XAuthLocation /opt/X11/bin/xauth
  ForwardAgent yes
  ForwardX11 yes
  SendEnv TERM=xterm-256color

  Setup SSH password:

  # remove old key
  ssh-keygen -R [localhost]:23456  # or remove manually from ~/.ssh/known_hosts
  ssh-copy-id vnpy # setup passwd

  Connect to VNPY container:

  In host, open a shell terminal, and run:

  ssh -X vnpy
  

• Sinopac account

  There are 10 of test accounts PAPIUSER01~PAPIUSER10 (All of the account's password are 2222) which can be used to connect to Sinopac gateway.

  There are limited functions available the simulation environment, so it is suggested to use real account to test VNPY.

  Logging with real account, you could get the live tick and also your current orders.

  It if you want to place order, you have to provide the certificate additionly.

  i.e. In host, copy the certificate to the working folder:

  cp Sinopac.pfx $VNPY_ROOT/run

• After SSHed in container, execute the following commands and the VNPY application window will show up.

  In container:

  cd workspace/run
  python run.py

Logging:

After logged in with simulation account:

• For the basic usage of VNPY, you can refer to:
  • 在 VNPY 上交易台灣股票、期貨 or
  • VnTrader基本使用.

System

• Remote SSH with GUI
• Developing inside a Container

VNPY

• VnTrader基本使用
• 在 VNPY 上交易台灣股票、期貨
• vn.py量化交易接口

Sinopac

• 在 VNPY 上交易台灣股票、期貨
• 聊天室
• Shioaji User Guide
• MAC 總管 供MAC電腦 申請/更新/查詢憑證

• Install ta-lib failed. error message: note: expected ‘const TA_FuncHandle **’

It's due to gcc run out of memory, enlarge the RAM size in Docker's preference and rebuilt it.

• qt.qpa.xcb: could not connect to display

ssh -X vnpy

• convert BIG-5

iconv -f BIG-5 -t UTF-8 excd.csv

• warnings.warn("urllib3 ({}) or chardet ({}) doesn't match a supported "

Re-install requests to resolve:

pip install requests