Simulation Utilities

Utility functions for simulation setup, runs and handling data.

sim_utils.check_sim_dup(sim_id: str, sim_num: int) → bool

Checks if a simulation number exists in the HDF5 file for a given simulation ID.

Parameters:

• sim_id (int) – The ID of the simulation.

• sim_num (int) – The number of the simulation to check for duplication.

Returns:

True if the simulation number exists in the file, False otherwise.

Return type:

bool

sim_utils.find_sim_num(params: dict, param_check: dict) → dict

Find simulation numbers that was run with a given set of paramters.

Useful for analysis. For e.g Finding the simulation number that was run with si_peak=1.0

Parameters:

• params (dict or Param) – Dictionary of all simulations generated after a run of multi simulation.

• param_check (dict) – A dictionary of parameters to check for. E.g: {si_peak: 1}

Returns:

A subset of global dictoinary containing only the simulations

that match the given parameters in param_check

Return type:

matches (dict)

sim_utils.get_git_commit_hash()

Retrieves the current Git commit hash.

Returns:

The Git commit hash as a string if the command is successful. None: If there is an error executing the command.

Return type:

str

sim_utils.get_module_from_path(file_path: str) → str

Convert a file path to a module name.

Used to import specs file that is passed as as argument to the simulation setup scripts.

Parameters:

file_path (str) – The relative file path (e.g., “specs/s_template.py”)

Returns:

The corresponding module name (e.g., “specs.s_template”)

Return type:

str

sim_utils.get_multiples_with_remainder(N, k)

Returns a list of multiples of k upto N.

Parameters:

• N (int) – The upper limit (exclusive) for the multiples.

• k (int) – The step size.

sim_utils.get_sim_num(iters: tuple, n_iters: tuple) → int

Calculate the simulation number for the running iterartor indices.

Useful for analysis.

Parameters:

• iters (tuple) – A tuple of current iteration indices.

• n_iters (tuple) – A tuple of the total size for each iterator.

Returns:

simulation number matching the given iterator indices.

Return type:

int

sim_utils.json_modify(obj: dict, fname: str)

Modify an entry in JSON file using the given dictionary.

If the file exists, it updates the file with the key-value pairs from the object. If the file does not exist, it creates a new file with the given object.

Parameters:

• obj (dict) – The dictionary object containing key-value pairs to be added or updated in the JSON file.

• fname (str) – The file name (including path) of the JSON file to be modified.

sim_utils.json_read(fname: str) → dict

Wrapper function to read a JSON file and return the dictionary object.

Parameters:

fname (str) – The file name to read the dictionary from.

Returns:

The dictionary object read from the JSON file.

Return type:

dict

sim_utils.json_save(obj: dict, fname: str)

Wrapper function to save a dictionary object to a JSON file.

Parameters:

• obj (dict) – The dictionary object to save.

• fname (str) – The file name to save the dictionary to.

sim_utils.list_to_numpy(LoL: list, fill: float = nan) → ndarray

Converts a list of lists (LoL) to a NumPy array, filling missing values with a specified fill value.

Used for to convert list spike times with non-homogeneous lengths to a NumPy array to save in hdf5 format.

Parameters:

• LoL (list of lists) – The input list of lists to be converted to a NumPy array.

• fill (float, optional) – The value to use for filling missing values. Defaults to np.nan.

Returns

numpy.ndarray: A NumPy array with the contents of the input list of lists, with missing values filled.

sim_utils.load_data(sim_id: str, data_id: str, cell_n: int = 0, sim_num: str = 0) → ndarray

Load Non-spiking data for a given simulation ID.

Parameters:

• sim_id (str) – Simulation ID to load data from.

• data_id (str) – Data ID. e.g. ‘stell_v’

• cell_n (int, optional) – Cell number to load data for, by default 0.

• sim_num (str, optional) – Simulation number to load data from, by default 0.

Returns:

The data array loaded from the HDF5 file

Return type:

np.ndarray

sim_utils.load_full_data(sim_id, data_id, sim_num=0)

Load Non-spiking data of all cells for a given simulation ID and number.

Parameters:

• sim_id (str) – Simulation ID to load data from.

• data_id (str) – Data ID. e.g. ‘stell_v’

• cell_n (int, optional) – Cell number to load data for, by default 0.

• sim_num (str, optional) – Simulation number to load data from, by default 0.

Returns:

The data array loaded from the HDF5 file.

Return type:

np.ndarray

sim_utils.load_sim_params(sim_id: str, file_path: str = None) → dict

Load simulation parameters.

Parameters:

• sim_id (str) – Simulation ID to load the parameters for.

• file_path (str, optional – Direct path to the JSON file containing the simulation parameters. If not provided, the parameters are located for the given sim_id.

Returns:

The simulation parameters loaded from the JSON file.

Return type:

dict

sim_utils.load_spikes(sim_id: str, sim_num: int = 0) → tuple

Load spike data for a given simulation ID.

This function looks for data in the following directories adn returns both stellate and interneuron spikes.

1. “data/” (local)

2. “/data/{user}/data/” (global)

Parameters:

• sim_id (str) – Simulation ID to load spikes from.

• sim_num (int, optional) – Simulation number to load spikes from, by default 0 (single sim).

Returns:

tuple

A tuple containing two lists of lists: - stell_spikes_l: List of spike times for stellate cells. - intrnrn_spikes_l: List of spike times for interneurons.

sim_utils.locate_data_dir(sim_id: str) → str

Find the location of the data directory for a given simulation ID.

The function checks for the existence of the data directory in several predefined locations.

1. “data/” (local)

2. “/data/{user}/data/” (global)

The {user} placeholder is replaced with the current user’s username obtained from the environment variables “USERNAME” or “USER”.

Parameters:

sim_id (str) – The sim ID to locate the data directory for.

Returns:

The location of the data directory.

Return type:

str

sim_utils.log_from_rank_0(logger: Logger, rank: int, msg: str, level: int = 20)

Logs a message if the rank is 0.

Parameters:

• logger (logging.Logger) – The logger object to use for logging.

• rank (int) – The rank of the process.

• msg (str) – The message to be logged.

sim_utils.network_intialize(params)

Initialize network and setup instrumentation

Parameters:

params (dict or Param) – Parameter dictionary

Returns:

Network object that includes cells and instrumentations

Return type:

network

sim_utils.process_data_root(data_root: str) → str

Processes the given data root path to ensure it ends with a slash.

Parameters:

data_root (str) – The root path to the data directory.

Returns:

The data root path ending with a slash.

Return type:

str

sim_utils.remove_nodes_from_params(params)

Remove “-node*” suffix from sim_id in multi params dict.

Parameters:

params (dict) – Multi params dict.

Returns:

modified params dict without “-node*” suffix from sim_id.

Return type:

dict

sim_utils.sim_run_arg_parser() → ArgumentParser

Set up the argument parser for the simulation run.

Argument parser is initialzed and processed here to avoid cluttering the main run file.

Returns:

The argument parser with the arguments added.

Return type:

argparse.ArgumentParser

sim_utils.sim_setup_arg_parser() → ArgumentParser

Set up the argument parser for the simulation.

Argument parser is initialzed and processed here to avoid cluttering the main setup file.

Returns:

The argument parser with the arguments added.

Return type:

argparse.ArgumentParser