: Implementing MCMC algorithms to find parameters in Return Dynamics
- "setup.txt" in the main directory is the setting file for how to implement MCMC
- It contains the options that choose either performing simulation or using observed data,
MCMC initial guess for paramters and hyperparameters
- Using the shell script below, you may implement MCMC and have figures that summarize final results
[ Summary for the parameters from MCMC implementations ]
![table](./plot_script/posterior_output.png)
[ Comparison between the true and estimated mu ]
![mu_last](./plot_script/mu_last.png)
The last 100 iteration-averaged mu
![mu_mean](./plot_script/mu_mean.png)
[ Chainplot for each parameter in the total iterations ]
![E_mu](./plot_script/E_mu_total.png)
![beta](./plot_script/beta_total.png)
![sigma_y](./plot_script/sigma_y_total.png)
![sigma_mu](./plot_script/sigma_mu_total.png)
![rho](./plot_script/rho_total.png)
[ Chainplot for each parameter in early stage of iterations ]
![E_mu_f](./plot_script/E_mu_first.png)
![beta_f](./plot_script/beta_first.png)
![sigma_y_f](./plot_script/sigma_y_first.png)
![sigma_mu_f](./plot_script/sigma_mu_first.png)
![rho_f](./plot_script/rho_first.png)
[ Prerequisites for this code in Ubuntu ]
Core library needed for C++
:~$ sudo apt-get install -y apt-get install libeigen3-dev
Graphic tool in Python script
sudo apt-get install -y python3 python3-pip python3-dev python3-env
sudo pip install numpy scipy
sudo pip install matplotlib