This is a cmake project to predict the number of infected persons by 2019-nCoV in China.
I'm trying to use a Susceptible-Infective Model to predict the number. This branch estimates the parameters of the formular using G2O with the data from Medsci.cn.
This is not a rigorous method to predict the number of infected persons, so don't take it seriously. As the evaluation criterion of definite diagnosis had changed since 2020 February 13th, I stopped updating new data for prediction.
I restarted updating at 2020 March 13th, and stopped updating at the end of March since the curve was approximately flat at that time.
G2O
For ROS Melodic or Noetic users, you can simply install it from apt:
$ sudo apt install ros-melodic-libg2o or
$ sudo apt install ros-noetic-libg2o
OpenCV
I have tested it with OpenCV 2.4.9, 3.2.0 and 4.2.0:
$ sudo apt install libopencv-dev
$ cd _2019_nCoV_infection_prediction
$ mkdir build
$ cd build
$ cmake ..
$ make
$ ./_2019_nCoV_infection_prediction