fixed sbi deprecation error. fixes in docs formatting

doc/source/user_documentation.rst

@@ -2710,6 +2710,7 @@ The ``update`` method will update the ``SimConfig`` object ``first`` with values
 This replaces the previous idiom for updating the SimConfig object with mappings from the batched job submission
 
 .. code-block:: python
+
     try:
         from __main__ import cfg  # import SimConfig object with params from parent module
     except:

netpyne/batch/batch.py

@@ -62,37 +62,40 @@ def tupleToStr(obj):
 class Batch(object):
     """
     Class that handles batch simulations on NetPyNE.
-    Relevant Attributes:
-        batchLabel : str
-            The label of the batch used for directory/file naming of batch generated files.
-        cfgFile : str
-            The path of the file containing the `netpyne.simConfig.SimConfig` object
-        cfg : `netpyne.simConfig.SimConfig`
-            The `netpyne.simConfig.SimConfig` object
-        N.B. either cfg or cfgFile should be specified #TODO: replace with typechecked single argument
-        netParamsFile : str
-            The path of the file containing the `netpyne.netParams.NetParams` object
-        netParams : `netpyne.netParams.NetParams`
-            The `netpyne.netParams.NetParams` object
-        N.B. either netParams or netParamsFile should be specified #TODO: replace with typechecked single argument
-        initCfg : dict
-            params dictionary that is used to modify the batch cfg prior to any algorithm based parameter modifications
-        saveFolder : str
-            The path of the folder where the batch will be saved (defaults to batchLabel)
-        method : str
-            The algorithm method used for batch
-        runCfg : dict
-            Keyword: Arg dictionary used to generate submission templates (see utils.py)
-        evolCfg : dict #TODO: replace with algoCfg? to merge with optimCfg
-            Keyword: Arg dictionary used to define evolutionary algorithm parameters (see evol.py)
-        optimCfg : dict #TODO: replace with algoCfg? to merge with evolCfg
-            Keyword: Arg dictionary used to define optimization algorithm parameters
-            (see asd_parallel.py, optuna_parallel.py, sbi_parallel.py)
-        params : list
-            Dictionary of parameters to be explored per algorithm (grid, evol, asd, optuna, sbi)
-            (see relevant algorithm script for details)
-        seed : int
-            Seed for random number generator for some algorithms
+
+    Attributes
+    ----------
+
+    batchLabel : str
+        The label of the batch used for directory/file naming of batch generated files.
+    cfgFile : str
+        The path of the file containing the `netpyne.simConfig.SimConfig` object
+    cfg : `netpyne.simConfig.SimConfig`
+        The `netpyne.simConfig.SimConfig` object
+    N.B. either cfg or cfgFile should be specified #TODO: replace with typechecked single argument
+    netParamsFile : str
+        The path of the file containing the `netpyne.netParams.NetParams` object
+    netParams : `netpyne.netParams.NetParams`
+        The `netpyne.netParams.NetParams` object
+    N.B. either netParams or netParamsFile should be specified #TODO: replace with typechecked single argument
+    initCfg : dict
+        params dictionary that is used to modify the batch cfg prior to any algorithm based parameter modifications
+    saveFolder : str
+        The path of the folder where the batch will be saved (defaults to batchLabel)
+    method : str
+        The algorithm method used for batch
+    runCfg : dict
+        Keyword: Arg dictionary used to generate submission templates (see utils.py)
+    evolCfg : dict #TODO: replace with algoCfg? to merge with optimCfg
+        Keyword: Arg dictionary used to define evolutionary algorithm parameters (see evol.py)
+    optimCfg : dict #TODO: replace with algoCfg? to merge with evolCfg
+        Keyword: Arg dictionary used to define optimization algorithm parameters
+        (see asd_parallel.py, optuna_parallel.py, sbi_parallel.py)
+    params : list
+        Dictionary of parameters to be explored per algorithm (grid, evol, asd, optuna, sbi)
+        (see relevant algorithm script for details)
+    seed : int
+        Seed for random number generator for some algorithms
     """
 
     def __init__(

netpyne/batch/sbi_parallel.py

@@ -20,16 +20,11 @@
 from neuron import h
 from sbi import utils as utils
 from sbi import analysis as analysis
-from sbi.inference.base import infer
-from sbi.analysis.plot import pairplot
 from sbi import utils
 from sbi import analysis
-from sbi import inference
-from sbi.inference import SNPE, SNLE, SNRE, simulate_for_sbi, prepare_for_sbi
+from sbi.inference import SNPE, SNLE, SNRE, simulate_for_sbi
+from sbi.inference.trainers.base import process_simulator, process_prior
 import torch
-import torch.nn as nn
-import torch.nn.functional as F
-from scipy.stats import kurtosis
 
 pc = h.ParallelContext()  # use bulletin board master/slave
 
@@ -188,7 +183,11 @@ def objectiveFunc(param):
             updateBestFit(candidate, fitness)
             return torch.as_tensor(fitness + sum_statistics)
 
-        simulator, prior = prepare_for_sbi(lambda param: objectiveFunc(param), prior)
+        # Check prior, return PyTorch prior.
+        prior, _, prior_returns_numpy = process_prior(prior)
+        # Check simulator, returns PyTorch simulator able to simulate batches.
+        simulator = process_simulator(simulator, prior, prior_returns_numpy)
+
         inference = sbi_md[sbi_method](prior=prior)
 
         if inference_type == 'single':

netpyne/batchtools/search.py

@@ -237,14 +237,19 @@ def generate_constructors(job_type, comm_type = 'socket', **kwargs):
 
 def generate_parameters(params, algorithm, **kwargs):
     """
-    returns a dictionary of parameters for ray_search based on the input dictionary
-    from NOTES Salvador:
+
+    Parameters
+    ----------
     params = {'synMechTau2': [3.0, 5.0, 7.0], # assumes list of values by default if grid search-like algo
 		  #'synMechTau2': [3.0, 7.0], # assumes lower/upper bounds by default if evol-like algo
           'connWeight' : paramtypes.sample_from(lambda _: numpy.random.uniform(0.005, 0.15))} # can optionally pass any of the paramtypes (= ray.tune data types)
 
-    #TODO: bloated function, prone to error
+    Returns
+    -------
+    a dictionary of parameters for ray_search based on the input dictionary
+
     """
+     #TODO: bloated function, prone to error
     ray_params = {}
     for param, space in params.items():
         if   isinstance(space, (list, tuple, range, numpy.ndarray)) and algorithm in {'variant_generator'}:

netpyne/cell/compartCell.py

@@ -1678,7 +1678,7 @@ def calcAbsSegCoords(self):
         self._segCoords['d1'] = morphSegCoords['d1']
 
     def setImembPtr(self):
-        """Set PtrVector to point to the i_membrane_"""
+        """Set PtrVector to point to the `i_membrane_`"""
         jseg = 0
         for sec in list(self.secs.values()):
             hSec = sec['hObj']

netpyne/network/stim.py

@@ -18,19 +18,19 @@ def addStims(self):
     """
     Internal function to add stims specified in specs.NetParams 
     Usage:
-        Creates and attaches stims to targets via CompartCell.addstim() based on entries in the specs.NetParams sub-
-        dictionaries -- specs.NetParams.stimSourceParams and specs.NetParams.stimTargetParams (see below)
-        NetParams.stimSourceParams entries contain key-value pairs to describe NEURON point processes specified by the
-        'type' entry (i.e. 'IClamp', 'VClamp', 'SEClamp', 'AlphaSynapse', 'VecStim')
-        NetParams.stimTargetParams entries contain key-value pairs to describe the post-synaptic connections for a
-        stimSourceParam entry specified by the 'source' entry, including a 'sec' and 'loc' entry (describing section
-        and location) for where the post-synaptic connection will exist and a 'conds' entry with a dictionary
-        specifying the cell criteria for the post-synaptic connections: (i.e. 'x', 'y', 'z' or 'xnorm', 'ynorm', 'znorm'
-        specifying cell criteria by location, 'cellList' specifying cell criteria by specific gid, or arbitrary
-        'key': 'value' tags.
-        For 'VecStim' point processes, it may be more convenient to create an artificial cell (i.e.netParams.popParams
-        see: netpyne/cell/pointCell.py) which allows pattern generation ('rhythmic', 'evoked', 'poisson', 'gauss')
-        by key-value entries in a 'spikePattern' dictionary.
+    Creates and attaches stims to targets via CompartCell.addstim() based on entries in the specs.NetParams sub-
+    dictionaries -- specs.NetParams.stimSourceParams and specs.NetParams.stimTargetParams (see below)
+    NetParams.stimSourceParams entries contain key-value pairs to describe NEURON point processes specified by the
+    'type' entry (i.e. 'IClamp', 'VClamp', 'SEClamp', 'AlphaSynapse', 'VecStim')
+    NetParams.stimTargetParams entries contain key-value pairs to describe the post-synaptic connections for a
+    stimSourceParam entry specified by the 'source' entry, including a 'sec' and 'loc' entry (describing section
+    and location) for where the post-synaptic connection will exist and a 'conds' entry with a dictionary
+    specifying the cell criteria for the post-synaptic connections: (i.e. 'x', 'y', 'z' or 'xnorm', 'ynorm', 'znorm'
+    specifying cell criteria by location, 'cellList' specifying cell criteria by specific gid, or arbitrary
+    'key': 'value' tags.
+    For 'VecStim' point processes, it may be more convenient to create an artificial cell (i.e.netParams.popParams
+    see: netpyne/cell/pointCell.py) which allows pattern generation ('rhythmic', 'evoked', 'poisson', 'gauss')
+    by key-value entries in a 'spikePattern' dictionary.
 
     Parameters
     ----------