README.txt

## TP IV
Main project : Learning the ising partition function 
For this, open Learning_Ising.ipynb and just follow the notebook !
It allows to learn the partition function of any hamiltonian on a spin up/down hilbert space with any number of particles, as long as there are interactions between spins and an external field. 


Explaination of different files

Notebooks : 
- Auto_regressive_flows.ipynb : Process for learning toy-models
- Diffusion_model.ipynb : Attempt at learning toy models using a diffusion model : currently not working
- Ising_partition_function_MCMC.ipynb : Visualize a Ising model and test its HST 
- Learning_Ising.ipynb : Final working algorithm to learn the partition function


Classes : 
- Models:- MeanField.py (Mean field approximation psi(s) = exp(sum bi si)
	 - Jastrow.py (Jastrow model psi(s) = exp (sum si Wij sj) )
- DiscreteOperator : contains the hamiltonians
	- IsingTransverse H = -J sum Z_i Z_i+1 -h sum X_i
	- X,Y,Z : Pauli matrices

- MCMC (engine to run MCMC loops)


Tests of classes : 
- Test_Models : testing the implementation of the models


Analysis : 
- VMC_<class>.py (VMC using MeanField or Jastrow model and MCMC)

Results :
- *.png : graphs of different simulations (all info in name + title)

Folders : 
- Ising_Analytical : graphs on the exact analytical results of the Ising model in 1D
- ClassicalCode : Code and results of the VMC on MeanField and Jastrow model
		  The Jastrow partition function is only correct for N_spins <= 3
- Angles_canceled : project about continuous space of angles : won't continue
- VMC : Codes for quantum MCMC (Semester 1)
- Graphes_and_Data : graphics and data of the converged solutions for the Ising HST