working on tutorial

README.md

@@ -55,7 +55,7 @@ where `NAME` need to be replaced by
 * `rc` for the RC model (Ex 3).
 * `rcr` for the RCR model (Ex 4).
 * `adaann` for the Friedman model example (Ex 5).
-* `rcr_nofas_adaann_example` for the RCR model, combining NoFAS with adaptive annealing (AdaAnn)
+* `rcr_nofas_adaann` for the RCR model, combining NoFAS with adaptive annealing (AdaAnn)
 
 At regular intervals set by the parameter `experiment.save_interval` LINFA writes a few results files. The sub-string `NAME` refers to the experiment name specified in the `experiment.name` variable, and `IT` indicates the iteration at which the file is written. The results files are
 
@@ -73,7 +73,7 @@ At regular intervals set by the parameter `experiment.save_interval` LINFA write
 A post processing script is also available to plot all results. To run it type
 
 ```sh
-python linfa.plot_res -n NAME -i IT -f FOLDER
+python -m linfa.plot_res -n NAME -i IT -f FOLDER
 ```
 where `NAME` and `IT` are again the experiment name and iteration number corresponding to the result file of interest, while `FOLDER` is the name of the folder with the results of the inference task are kept.
 

docs/content/rcr.rst

@@ -1,7 +1,7 @@
 Three-element Wndkessel Model
 =============================
 
-The three-parameter Windkessel or **RCR** model is characterized by proximal and distal resistance parameters :math:`R_{p}, R_{d} \in [100, 1500]`$ Barye :math:`\cdot` s/ml and one capacitance parameter :math:`C \in [1\times 10^{-5}, 1\times 10^{-2}]` ml/Barye.
+The three-parameter Windkessel or **RCR** model is characterized by proximal and distal resistance parameters :math:`R_{p}, R_{d} \in [100, 1500]` Barye :math:`\cdot` s/ml and one capacitance parameter :math:`C \in [1\times 10^{-5}, 1\times 10^{-2}]` ml/Barye.
 
 This model is not identifiable. The average distal pressure is only affected by the total system resistance, i.e. the sum :math:`R_{p}+R_{d}`, leading to a negative correlation between these two parameters. Thus, an increment in the proximal resistance is compensated by a reduction in the distal resistance (so the average distal pressure remains the same) which, in turn, reduces the friction encountered by the flow exiting the capacitor. An increase in the value of :math:`C` is finally needed to restore the average, minimum and maximum pressure. This leads to a positive correlation between :math:`C` and :math:`R_{d}`.
 
@@ -14,14 +14,14 @@ where the distal pressure is set to :math:`P_{d}=55` mmHg.
 
 Synthetic observations are generated from :math:`N(\boldsymbol\mu, \boldsymbol\Sigma)`, where :math:`\mu=(f_{1}(\boldsymbol{z}^{*}),f_{2}(\boldsymbol{z}^{*}),f_{3}(\boldsymbol{z}^{*}))^T = (P_{p,\text{min}}, P_{p,\text{max}}, P_{p,\text{ave}})^T = (100.96, 148.02,116.50)^T` and :math:`\boldsymbol\Sigma`` is a diagonal matrix with entries :math:`(5.05, 7.40, 5.83)^T`. The budgeted number of true model solutions is 216; the fixed surrogate model is evaluated on a :math:`6\times 6\times 6 = 216` pre-grid while the adaptive surrogate is evaluated with a pre-grid of size :math:`4\times 4\times 4 = 64` and the other 152 evaluations are adaptively selected.
 
-The results are presented in :numref:`fig_rcr_res`. The posterior samples obtained through NoFAS capture well the non-linear correlation among the parameters and generate a fairly accurate posterior predictive distribution that overlaps with the observations. Additional details can be found in :cite:p:`wang2022variational`.
+This example also demonstrates how NoFAS can be combined with annealing for improved convergence. The results in :numref:`fig_rcr_res` are generated using the AdaAnn adaptive annealing scheduler with intial inverse temperature :math:`t_{0}=0.05`, KL tolerance :math:`\tau=0.01` and a batch size of 100 samples. The number of parameter updates is set to 500, 5000 and 5 for :math:`t_{0}`, :math:`t_{1}` and :math:`t_{0}<t<t_{1}`, respectively and 1000 Monte Carlo realizations are used to evaluate the denominator in :eq:`eqn:friedman1`. The posterior samples capture well the nonlinear correlation among the parameters and generate a fairly accurate posterior predictive distribution that overlaps with the observations. Additional details can be found in :cite:p:`wang2022variational,cobian2023adaann`.
 
-.. figure:: imgs/rcr/log_plot_rcr-1.png
-.. figure:: imgs/rcr/data_plot_rcr_25000_0_1-1.png
-.. figure:: imgs/rcr/data_plot_rcr_25000_0_2-1.png
-.. figure:: imgs/rcr/params_plot_rcr_25000_0_1-1.png
-.. figure:: imgs/rcr/params_plot_rcr_25000_0_2-1.png
-.. figure:: imgs/rcr/params_plot_rcr_25000_1_2-1.png
+.. figure:: imgs/rcr/log_plot-1.png
+.. figure:: imgs/rcr/data_plot_rcr_nofas_adaann_8400_0_1-1.png
+.. figure:: imgs/rcr/data_plot_rcr_nofas_adaann_8400_0_2-1.png
+.. figure:: imgs/rcr/params_plot_rcr_nofas_adaann_8400_0_1-1.png
+.. figure:: imgs/rcr/params_plot_rcr_nofas_adaann_8400_0_2-1.png
+.. figure:: imgs/rcr/params_plot_rcr_nofas_adaann_8400_1_2-1.png
    :name: fig_rcr_res
 
    Results from the RCR model. Loss profile (top), posterior predictive distribution (center) and posterior samples (bottom).

linfa/plot_res.py

@@ -81,7 +81,7 @@ def plot_outputs(sample_file,obs_file,idx1,idx2,out_dir,out_info):
                       action=None,
                       # nargs='+',
                       const=None,
-                      default='./tests/results/',
+                      default='./',
                       type=str,
                       required=False,
                       help='Folder with experiment results',
@@ -132,7 +132,7 @@ def plot_outputs(sample_file,obs_file,idx1,idx2,out_dir,out_info):
     print('Plotting log...')
     plot_log(log_file,out_dir)
   else:
-    print('Log file not found...')
+    print('Log file not found: '+log_file)
 
   # Plot 2D slice of posterior samples
   if(os.path.isfile(param_file) and os.path.isfile(LL_file)):
@@ -142,7 +142,8 @@ def plot_outputs(sample_file,obs_file,idx1,idx2,out_dir,out_info):
       for loopB in range(loopA+1, tot_params):
         plot_params(param_file,LL_file,loopA,loopB,out_dir,out_info)
   else:
-    print('File with posterior samples not found...')
+    print('File with posterior samples not found: '+param_file)
+    print('File with log-density not found: '+LL_file)
 
   # Plot 2D slice of outputs and observations
   if(os.path.isfile(output_file) and os.path.isfile(obs_file)):
@@ -152,4 +153,5 @@ def plot_outputs(sample_file,obs_file,idx1,idx2,out_dir,out_info):
       for loopB in range(loopA+1, tot_outputs):
         plot_outputs(output_file,obs_file,loopA,loopB,out_dir,out_info)
   else:
-    print('File with posterior predictive samples not found...')
+    print('File with posterior predictive samples not found: '+output_file)
+    print('File with observations not found: '+obs_file)

pyproject.toml

@@ -6,7 +6,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "linfa_vi"
-version = "1.1.5"
+version = "1.1.6"
 description = "A Python library for inference with normalizing flow and annealing"
 readme = "README.md"
 authors = [{ name = "resDesLab ", email = "daniele.schiavazzi@gmail.com" }]