Refactor and improve plotting of time series and spikes from simulation generated data

docs/source/pyneuroml.utils.rst

@@ -38,6 +38,14 @@ pyneuroml.utils.cli module
    :undoc-members:
    :show-inheritance:
 
+pyneuroml.utils.simdata module
+-------------------------------
+
+.. automodule:: pyneuroml.utils.simdata
+   :members:
+   :undoc-members:
+   :show-inheritance:
+
 pyneuroml.utils.units module
 ----------------------------
 

examples/LeakConductance.mod

@@ -3,9 +3,9 @@ TITLE Mod file for component: Component(id=LeakConductance type=ionChannelHH)
 COMMENT
 
     This NEURON file has been generated by org.neuroml.export (see https://github.com/NeuroML/org.neuroml.export)
-         org.neuroml.export  v1.10.1
-         org.neuroml.model   v1.10.1
-         jLEMS               v0.11.1
+         org.neuroml.export  v1.11.0
+         org.neuroml.model   v1.11.0
+         jLEMS               v0.12.0
 
 ENDCOMMENT
 

pyneuroml/archive/__init__.py

@@ -13,10 +13,10 @@
 import typing
 from zipfile import ZipFile
 
-from pyneuroml.runners import run_jneuroml
-from pyneuroml.sedml import validate_sedml_files
-from pyneuroml.utils import get_model_file_list
-from pyneuroml.utils.cli import build_namespace
+import pyneuroml.runners as pynmlr
+import pyneuroml.sedml as pynmls
+import pyneuroml.utils as pynmlu
+import pyneuroml.utils.cli as pynmluc
 from pyneuroml.utils.misc import chdir
 
 logger = logging.getLogger(__name__)
@@ -85,7 +85,7 @@ def main(args=None):
 
 def cli(a: typing.Optional[typing.Any] = None, **kwargs: str):
     """Main cli caller method"""
-    a = build_namespace(DEFAULTS, a, **kwargs)
+    a = pynmluc.build_namespace(DEFAULTS, a, **kwargs)
 
     rootfile = a.rootfile
     zipfile_extension = ".neux.zip"
@@ -96,13 +96,13 @@ def cli(a: typing.Optional[typing.Any] = None, **kwargs: str):
         a.rootfile.startswith("LEMS") and a.rootfile.endswith(".xml")
     ) and a.sedml is True:
         logger.debug("Generating SED-ML file from LEMS file")
-        run_jneuroml("", a.rootfile, "-sedml")
+        pynmlr.run_jneuroml("", a.rootfile, "-sedml")
 
         rootfile = a.rootfile.replace(".xml", ".sedml")
         zipfile_extension = ".omex.zip"
 
         # validate the generated file
-        validate_sedml_files([rootfile])
+        pynmls.validate_sedml_files([rootfile])
 
     # if explicitly given, use that
     if a.zipfile_extension is not None:
@@ -173,10 +173,9 @@ def create_combine_archive(
     with chdir(rootdir):
         lems_def_dir = None
         if len(filelist) == 0:
-            lems_def_dir = get_model_file_list(
+            lems_def_dir = pynmlu.get_model_file_list(
                 rootfile, filelist, rootdir, lems_def_dir
             )
-
         create_combine_archive_manifest(rootfile, filelist + extra_files, rootdir)
         filelist.append("manifest.xml")
 

pyneuroml/lems/LEMSSimulation.py

@@ -13,7 +13,7 @@
 from neuroml import __version__ as libnml_ver
 
 from pyneuroml import __version__ as pynml_ver
-from pyneuroml.pynml import read_lems_file, read_neuroml2_file
+from pyneuroml.io import read_lems_file, read_neuroml2_file
 from pyneuroml.utils.plot import get_next_hex_color
 from pyneuroml.utils.units import convert_to_units
 

pyneuroml/lems/__init__.py

@@ -6,10 +6,9 @@
 import typing
 
 import neuroml
-from lxml import etree
 
+from pyneuroml.io import read_neuroml2_file
 from pyneuroml.lems.LEMSSimulation import LEMSSimulation
-from pyneuroml.pynml import read_neuroml2_file
 from pyneuroml.utils.plot import get_next_hex_color
 
 logger = logging.getLogger(__name__)
@@ -480,184 +479,3 @@ def get_pop_index(quantity):
         pop = s[0]
         index = int(s[1])
         return pop, index
-
-
-def load_sim_data_from_lems_file(
-    lems_file_name: str,
-    base_dir: str = ".",
-    get_events: bool = True,
-    get_traces: bool = True,
-) -> typing.Optional[typing.Union[typing.Tuple[typing.Dict, typing.Dict], typing.Dict]]:
-    """Load simulation outputs using the LEMS simulation file
-
-    .. versionadded:: 1.2.2
-
-    :param lems_file_name: name of LEMS file that was used to generate the data
-    :type lems_file_name: str
-    :param base_dir: directory to run in
-    :type base_dir: str
-    :returns: if both `get_events` and `get_traces` are selected, a tuple with
-        two dictionaries, one for traces, one for events, is returned.
-
-        Otherwise one dictionary for whichever was selected.
-
-        The events dictionary has the following format:
-
-        .. code-block:: python
-
-            {
-                '<value of select attribute>': { 'cell id': [<events>] }
-            }
-
-        The traces dictionary has the following format:
-
-        .. code-block:: python
-
-            {
-                't': [<values>],
-                'col 1': [<values>]
-                'col 2': [<values>]
-            }
-
-    :raises ValueError: if neither traces nor events are selected for loading
-    :raises ValueError: if no traces are found
-    :raises ValueError: if no events are found
-
-    """
-    if not os.path.isfile(lems_file_name):
-        real_lems_file = os.path.realpath(os.path.join(base_dir, lems_file_name))
-    else:
-        real_lems_file = os.path.realpath(lems_file_name)
-
-    if not get_events and not get_traces:
-        raise ValueError("One of events or traces must be True")
-
-    logger.debug(
-        "Reloading data specified in LEMS file: %s (%s), base_dir: %s, cwd: %s;"
-        % (lems_file_name, real_lems_file, base_dir, os.getcwd())
-    )
-
-    # Could use pylems to parse all this...
-    traces = {}  # type: dict
-    events = {}  # type: dict
-
-    base_lems_file_path = os.path.dirname(os.path.realpath(lems_file_name))
-    tree = etree.parse(real_lems_file)
-
-    sim = tree.getroot().find("Simulation")
-    ns_prefix = ""
-
-    possible_prefixes = ["{http://www.neuroml.org/lems/0.7.2}"]
-    if sim is None:
-        for pre in possible_prefixes:
-            for comp in tree.getroot().findall(pre + "Component"):
-                if comp.attrib["type"] == "Simulation":
-                    ns_prefix = pre
-                    sim = comp
-
-    if get_events:
-        event_output_files = sim.findall(ns_prefix + "EventOutputFile")
-        for i, of in enumerate(event_output_files):
-            name = of.attrib["fileName"]
-            file_name = os.path.join(base_dir, name)
-            if not os.path.isfile(file_name):  # If not relative to the LEMS file...
-                file_name = os.path.join(base_lems_file_path, name)
-
-            # if not os.path.isfile(file_name): # If not relative to the LEMS file...
-            #    file_name = os.path.join(os.getcwd(),name)
-            # ... try relative to cwd.
-            # if not os.path.isfile(file_name): # If not relative to the LEMS file...
-            #    file_name = os.path.join(os.getcwd(),'NeuroML2','results',name)
-            # ... try relative to cwd in NeuroML2/results subdir.
-            if not os.path.isfile(file_name):  # If not relative to the base dir...
-                raise OSError(
-                    ("Could not find simulation output " "file %s" % file_name)
-                )
-            format_ = of.attrib["format"]
-            logger.info(
-                "Loading saved events from %s (format: %s)" % (file_name, format_)
-            )
-            selections = {}
-            for col in of.findall(ns_prefix + "EventSelection"):
-                id_ = int(col.attrib["id"])
-                select = col.attrib["select"]
-                events[select] = []
-                selections[id_] = select
-
-            with open(file_name) as f:
-                for line in f:
-                    values = line.split()
-                    if format_ == "TIME_ID":
-                        t = float(values[0])
-                        id_ = int(values[1])
-                    elif format_ == "ID_TIME":
-                        id_ = int(values[0])
-                        t = float(values[1])
-                    if id_ in selections:
-                        logger.debug(
-                            "Found a event in cell %s (%s) at t = %s"
-                            % (id_, selections[id_], t)
-                        )
-                        events[selections[id_]].append(t)
-
-                    else:
-                        logger.warning("ID %s not found in selections dictionary" % id_)
-                        continue  # skip this event
-
-    if get_traces:
-        output_files = sim.findall(ns_prefix + "OutputFile")
-
-        for i, of in enumerate(output_files):
-            traces["t"] = []
-            name = of.attrib["fileName"]
-            file_name = os.path.join(base_dir, name)
-
-            if not os.path.isfile(file_name):  # If not relative to the LEMS file...
-                file_name = os.path.join(base_lems_file_path, name)
-
-            if not os.path.isfile(file_name):  # If not relative to the LEMS file...
-                file_name = os.path.join(os.getcwd(), name)
-
-                # ... try relative to cwd.
-            if not os.path.isfile(file_name):  # If not relative to the LEMS file...
-                file_name = os.path.join(os.getcwd(), "NeuroML2", "results", name)
-                # ... try relative to cwd in NeuroML2/results subdir.
-            if not os.path.isfile(file_name):  # If not relative to the LEMS file...
-                raise OSError(
-                    ("Could not find simulation output " "file %s" % file_name)
-                )
-
-            logger.info("Loading traces from %s" % (file_name))
-            cols = []
-            cols.append("t")
-            for col in of.findall(ns_prefix + "OutputColumn"):
-                quantity = col.attrib["quantity"]
-                traces[quantity] = []
-                cols.append(quantity)
-
-            # TODO: could be quicker using numpy etc?
-            with open(file_name) as f:
-                for line in f:
-                    values = line.split()
-                    for vi in range(len(values)):
-                        traces[cols[vi]].append(float(values[vi]))
-
-    if get_events is True and get_traces is True:
-        if len(events) == 0:
-            raise ValueError("No events found")
-        if len(traces) == 0:
-            raise ValueError("No traces found")
-        logger.debug("Returning both traces and events")
-        return traces, events
-    else:
-        if get_events is True:
-            if len(events) == 0:
-                raise ValueError("No events found")
-            logger.debug("Returning events")
-            return events
-        elif get_traces is True:
-            if len(traces) == 0:
-                raise ValueError("No traces found")
-            logger.debug("Returning traces")
-            return traces
-    return None

pyneuroml/plot/Plot.py

@@ -140,6 +140,8 @@ def generate_plot(
 
     :type legend_position: str
     :param show_plot_already: if plot should be shown when created (default: True)
+        Otherwise, the plots are not shown and you must call pyplot.show()
+        explicitly to show them. Should not be used with `close_plot`.
     :type show_plot_already: boolean
     :param animate: if plot should be animated (default: False)
     :type animate: boolean
@@ -207,7 +209,7 @@ def generate_plot(
             f"values to plot ({len(xvalues)}) and markersizes ({len(markersizes)}) must have the same length"
         )
 
-    logger.info("Generating plot: %s" % (title))
+    logger.info("Generating plot" + (f": {title}" if title else ""))
 
     from matplotlib import pyplot as plt
     from matplotlib import rcParams
@@ -380,15 +382,11 @@ def update(frame):
                 )
 
             pbar.finish(dirty=False)
-            logger.info("Saved animation to %s" % (save_figure_to))
+            logger.info(f"Saved animation to {os.path.abspath(save_figure_to)}")
     else:
         if save_figure_to:
-            logger.info(
-                "Saving image to %s of plot: %s"
-                % (os.path.abspath(save_figure_to), title)
-            )
             plt.savefig(save_figure_to, bbox_inches="tight")
-            logger.info("Saved image to %s of plot: %s" % (save_figure_to, title))
+            logger.info(f"Saved image to {os.path.abspath(save_figure_to)}")
 
     if show_plot_already:
         if interactive_legend is True and legend_box is not None:

pyneuroml/plot/PlotSpikes.py

@@ -18,7 +18,7 @@
 import matplotlib.pyplot as plt
 import numpy as np
 
-import pyneuroml.lems as pynmll
+import pyneuroml.utils.simdata as pynmls
 from pyneuroml.plot import generate_plot
 from pyneuroml.utils.cli import build_namespace
 
@@ -480,7 +480,7 @@ def plot_spikes_from_lems_file(
     :type lems_file_name: str
     :param base_dir: Directory where the LEMS file resides. Defaults to the current directory.
     :type base_dir: str
-    :param show_plots_already: Whether to show the plots immediately after they are generated. Defaults to True.
+    :param show_plots_already: Whether to show the plots
     :type show_plots_already: bool
     :param save_spike_plot_to: Path to save the spike plot to. If `None`, the plot will not be saved. Defaults to `None`.
     :type save_spike_plot_to: Optional[str]
@@ -493,31 +493,38 @@ def plot_spikes_from_lems_file(
     :return: None
     :rtype: None
     """
-    event_data = pynmll.load_sim_data_from_lems_file(
+    all_events: Dict[str, Dict] = pynmls.load_sim_data_from_lems_file(
         lems_file_name, get_events=True, get_traces=False
     )
 
-    spike_data = []  # type: List[Dict]
-    for select, times in event_data.items():
-        new_dict = {"name": select}
-        new_dict["times"] = times
-        # the plot_spikes function will add an offset for each data entry, so
-        # we set the ids to 0 here
-        new_dict["ids"] = [0] * len(times)
-
-        spike_data.append(new_dict)
-
-    logger.debug("Spike data is:")
-    logger.debug(spike_data)
+    if len(all_events) > 1:
+        show_each_plot_already = False
+
+    for filename, event_data in all_events.items():
+        spike_data = []  # type: List[Dict]
+        for select, times in event_data.items():
+            new_dict = {"name": select}
+            new_dict["times"] = times
+            # the plot_spikes function will add an offset for each data entry, so
+            # we set the ids to 0 here
+            new_dict["ids"] = [0] * len(times)
+
+            spike_data.append(new_dict)
+
+        logger.debug("Spike data is:")
+        logger.debug(spike_data)
+
+        plot_spikes(
+            spike_data,
+            show_plots_already=show_each_plot_already,
+            save_spike_plot_to=save_spike_plot_to,
+            rates=rates,
+            rate_window=rate_window,
+            rate_bins=rate_bins,
+        )
 
-    plot_spikes(
-        spike_data,
-        show_plots_already=show_plots_already,
-        save_spike_plot_to=save_spike_plot_to,
-        rates=rates,
-        rate_window=rate_window,
-        rate_bins=rate_bins,
-    )
+    if show_plots_already is True and show_each_plot_already is False:
+        plt.show()
 
 
 def main(args: Optional[argparse.Namespace] = None) -> None: