Still working on the data buffer

setup.cfg

@@ -40,3 +40,4 @@ where = src
 [options.entry_points]
 console_scripts = 
     scgenerator = scgenerator.cli.cli:main
+    sc-submit = scgenerator.scripts.slurm_submit:main

src/scgenerator/cli/cli.py

@@ -35,8 +35,6 @@ def create_parser():
     )
     resume_parser.set_defaults(func=resume_sim)
 
-    newconfig_parser = subparsers.add_parser("newconfig", help="create a new configuration file")
-
     return parser
 
 

src/scgenerator/cli/new_config.py

@@ -1,21 +1,27 @@
 from .. import const
+import toml
+
+valid_commands = ["finish", "next"]
 
 
-def list_input():
+class Configurator:
+    def __init__(self, name):
+        self.config = dict(name=name, fiber=dict(), gas=dict(), pulse=dict(), simulation=dict())
+
+    def list_input(self):
         answer = ""
         while answer == "":
             answer = input("Please enter a list of values (one per line)\n")
 
-    out = [process_input(answer)]
+        out = [self.process_input(answer)]
 
         while answer != "":
             answer = input()
-        out.append(process_input(answer))
+            out.append(self.process_input(answer))
 
         return out[:-1]
 
-
-def process_input(s):
+    def process_input(self, s):
         try:
             return int(s)
         except ValueError:
@@ -28,8 +34,7 @@ def process_input(s):
 
         return s
 
-
-def accept(question, default=True):
+    def accept(self, question, default=True):
         question += " ([y]/n)" if default else " (y/[n])"
         question += "\n"
         inp = input(question)
@@ -40,8 +45,10 @@ def accept(question, default=True):
 
         return inp.lower() in yes_str
 
+    def print_current(self, config: dict):
+        print(toml.dumps(config))
 
-def get(section, param_name):
+    def get(self, section, param_name):
         question = f"Please enter a value for the parameter '{param_name}'\n"
         valid = const.valid_param_types[section][param_name]
 
@@ -50,12 +57,22 @@ def get(section, param_name):
 
         while not is_valid:
             answer = input(question)
-        if answer == "\\variable" and param_name in const.valid_variable[section]:
-            value = list_input()
+            if answer == "variable" and param_name in const.valid_variable[section]:
+                value = self.list_input()
                 print(value)
                 is_valid = all(valid(v) for v in value)
             else:
-            value = process_input(answer)
+                value = self.process_input(answer)
                 is_valid = valid(value)
 
         return value
+
+    def ask_next_command(self):
+        s = ""
+        raw_input = input(s).split(" ")
+        return raw_input[0], raw_input[1:]
+
+    def main(self):
+        editing = True
+        while editing:
+            command, args = self.ask_next_command()

src/scgenerator/const.py

@@ -1,6 +1,26 @@
 import numpy as np
 
 
+def in_range_excl(func, r):
+    def _in_range(n):
+        if not func(n):
+            return False
+        return n > r[0] and n < r[1]
+
+    _in_range.__doc__ = func.__doc__ + f" between {r[0]} and {r[1]}"
+    return _in_range
+
+
+def in_range_incl(func, r):
+    def _in_range(n):
+        if not func(n):
+            return False
+        return n >= r[0] and n <= r[1]
+
+    _in_range.__doc__ = func.__doc__ + f" between {r[0]} and {r[1]}"
+    return _in_range
+
+
 def num(n):
     """must be a single, real, non-negative number"""
     return isinstance(n, (float, int)) and n >= 0
@@ -83,32 +103,24 @@ def capillary_nested(n):
     return isinstance(n, int) and n >= 0
 
 
-# def find_parent(param):
-#     """find the parent dictionary name of param"""
-#     for k, v in valid_param_types.items():
-#         if param in v:
-#             return k
-#     raise ValueError(f"'{param}' is an invalid parameter name")
-
-
 valid_param_types = dict(
     root=dict(
         name=lambda s: isinstance(s, str),
     ),
     fiber=dict(
         gamma=num,
-        pitch=num,
-        pitch_ratio=num,
-        core_radius=num,
+        pitch=in_range_excl(num, (0, 1e-3)),
+        pitch_ratio=in_range_excl(num, (0, 1)),
+        core_radius=in_range_excl(num, (0, 1e-3)),
         he_mode=he_mode,
         fit_parameters=fit_parameters,
         beta=beta,
         model=string(["pcf", "marcatili", "marcatili_adjusted", "hasan", "custom"]),
         length=num,
         capillary_num=integer,
-        capillary_outer_d=num,
-        capillary_thickness=num,
-        capillary_spacing=num,
+        capillary_outer_d=in_range_excl(num, (0, 1e-3)),
+        capillary_thickness=in_range_excl(num, (0, 1e-3)),
+        capillary_spacing=in_range_excl(num, (0, 1e-3)),
         capillary_resonance_strengths=capillary_resonance_strengths,
         capillary_nested=capillary_nested,
     ),
@@ -125,10 +137,10 @@ valid_param_types = dict(
         soliton_num=num,
         quantum_noise=boolean,
         shape=string(["gaussian", "sech"]),
-        wavelength=num,
-        intensity_noise=num,
-        width=num,
-        t0=num,
+        wavelength=in_range_excl(num, (100e-9, 3000e-9)),
+        intensity_noise=in_range_incl(num, (0, 1)),
+        width=in_range_excl(num, (0, 1e-9)),
+        t0=in_range_excl(num, (0, 1e-9)),
     ),
     simulation=dict(
         behaviors=behaviors,
@@ -139,11 +151,11 @@ valid_param_types = dict(
         t_num=integer,
         z_num=integer,
         time_window=num,
-        dt=num,
-        tolerated_error=num,
+        dt=in_range_excl(num, (0, 5e-15)),
+        tolerated_error=in_range_excl(num, (1e-15, 1e-5)),
         step_size=num,
-        lower_wavelength_interp_limit=num,
-        upper_wavelength_interp_limit=num,
+        lower_wavelength_interp_limit=in_range_excl(num, (100e-9, 3000e-9)),
+        upper_wavelength_interp_limit=in_range_excl(num, (100e-9, 5000e-9)),
         frep=num,
     ),
 )

src/scgenerator/data/start_head.sh

@@ -0,0 +1,9 @@
+#!/bin/bash
+
+echo "starting ray head node"
+
+# Launch the head node
+ray start --head --node-ip-address=$1 --port=6379 --redis-password=$2 --num-cpus=$SLURM_CPUS_PER_TASK --dashboard-host=127.0.0.1
+
+
+sleep infinity

src/scgenerator/data/start_worker.sh

@@ -0,0 +1,5 @@
+#!/bin/bash
+
+echo "starting ray worker node"
+ray start --address $1 --redis-password=$2 --num-cpus=$SLURM_CPUS_PER_TASK
+sleep infinity

src/scgenerator/data/submit_job_template.txt

@@ -0,0 +1,52 @@
+#!/bin/bash
+
+#SBATCH --time={time}
+
+### This script works for any number of nodes, Ray will find and manage all resources
+#SBATCH --nodes={nodes}
+
+### Give all resources to a single Ray task, ray can manage the resources internally
+#SBATCH --ntasks-per-node=1
+#SBATCH --cpus-per-task={cpus_per_node}
+
+
+# Load modules or your own conda environment here
+{environment_setup}
+
+################# DO NOT CHANGE THINGS HERE UNLESS YOU KNOW WHAT YOU ARE DOING ###############
+# This script is a modification to the implementation suggest by gregSchwartz18 here:
+# https://github.com/ray-project/ray/issues/826#issuecomment-522116599
+redis_password=$(uuidgen)
+export redis_password
+
+nodes=$(scontrol show hostnames $SLURM_JOB_NODELIST) # Getting the node names
+nodes_array=( $nodes )
+
+node_1=${{nodes_array[0]}} 
+ip=$(srun --nodes=1 --ntasks=1 -w $node_1 hostname --ip-address) # making redis-address
+echo $ip > ~/ip_head
+port=6379
+ip_head=$ip:$port
+export ip_head
+echo "IP Head: $ip_head"
+
+ip_submit=$(hostname --ip-address)
+
+echo "STARTING HEAD at $node_1"
+srun --nodes=1 --ntasks=1 -w $node_1 start_head.sh $ip $redis_password &
+sleep 5
+
+
+worker_num=$(($SLURM_JOB_NUM_NODES - 1)) #number of nodes other than the head node
+for ((  i=1; i<=$worker_num; i++ ))
+do
+  node_i=${{nodes_array[$i]}}
+  echo "STARTING WORKER $i at $node_i"
+  srun --nodes=1 --ntasks=1 -w $node_i start_worker.sh $ip_head $redis_password &
+  sleep 5
+done
+##############################################################################################
+
+#### call your code below
+scgenerator run {config}
+exit

src/scgenerator/initialize.py

@@ -153,7 +153,7 @@ def validate_single_parameter(section, key, value):
         raise TypeError(s)
     if not func(value):
         raise TypeError(
-            f"value '{value}' of type {type(value)} for key '{key}' is not valid, {func.__doc__}"
+            f"value '{value}' of type {type(value).__name__} for key '{key}' is not valid, {func.__doc__}"
         )
     return
 

src/scgenerator/io.py

@@ -26,8 +26,14 @@ except ModuleNotFoundError:
 
 
 class Paths:
-    home = os.path.expanduser("~")
-    _data_files = ["silica.toml", "gas.toml", "hr_t.npz"]
+    _data_files = [
+        "silica.toml",
+        "gas.toml",
+        "hr_t.npz",
+        "submit_job_template.txt",
+        "start_worker.sh",
+        "start_head.sh",
+    ]
 
     paths = {
         f.split(".")[0]: os.path.abspath(
@@ -81,6 +87,8 @@ class DataBuffer:
 
     def empty(self):
         num = self.queue.size()
+        if num == 0:
+            return 0
         self.logger.info(f"buffer length at time of emptying : {num}")
         while not self.queue.empty():
             name, identifier, data = self.queue.get()
@@ -372,8 +380,8 @@ def merge_same_simulations(path: str):
         if len(base_folder) > 0:
             base_folders.add(base_folder)
 
-    num_operations = z_num * len(base_folders) + len(base_folders)
-    pt = utils.ProgressTracker(num_operations, logger=logger, prefix="merging data : ")
+    sim_num, param_num = utils.count_variations(config)
+    pt = utils.ProgressTracker(sim_num, logger=logger, prefix="merging data : ")
 
     spectra = []
     for z_id in range(z_num):
@@ -386,6 +394,7 @@ def merge_same_simulations(path: str):
 
             in_path = os.path.join(path, utils.format_variable_list(variable_and_ind))
             spectra.append(np.load(os.path.join(in_path, f"spectrum_{z_id}.npy")))
+            pt.update()
 
             # write new files only once all those from one parameter set are collected
             if repeat_id == max_repeat_id:
@@ -393,7 +402,6 @@ def merge_same_simulations(path: str):
                 out_path = ensure_folder(out_path, prevent_overwrite=False)
                 spectra = np.array(spectra).reshape(repeat, len(spectra[0]))
                 np.save(os.path.join(out_path, f"spectra_{z_id}.npy"), spectra.squeeze())
-                pt.update()
 
                 # copy other files only once
                 if z_id == 0:
@@ -402,7 +410,6 @@ def merge_same_simulations(path: str):
                             os.path.join(in_path, file_name),
                             os.path.join(out_path, ""),
                         )
-                    pt.update()
 
     try:
         for sub_folder in sub_folders:

src/scgenerator/logger.py

@@ -49,7 +49,7 @@ def get_logger(name=None):
 #             handler.setLevel(lvl_map[lvl])
 
 
-def configure_logger(logger, logfile="scgenerator.log"):
+def configure_logger(logger):
     """configures a logging.Logger obj
 
     Parameters
@@ -65,16 +65,16 @@ def configure_logger(logger, logfile="scgenerator.log"):
         updated logger
     """
     if not hasattr(logger, "already_configured"):
-        if logfile is not None:
-            file_handler = logging.FileHandler("scgenerator.log", "a+")
-            file_handler.setFormatter(
-                logging.Formatter("{levelname}: {name}: {message}", style="{")
-            )
-            logger.addHandler(file_handler)
+        formatter = logging.Formatter("{levelname}: {name}: {message}", style="{")
+        file_handler1 = logging.FileHandler("sc-DEBUG.log", "a+")
+        file_handler1.setFormatter(formatter)
+        file_handler1.setLevel(logging.DEBUG)
+        logger.addHandler(file_handler1)
 
         stream_handler = logging.StreamHandler()
+        stream_handler.setLevel(logging.INFO)
         logger.addHandler(stream_handler)
-        logger.setLevel(DEFAULT_LEVEL)
+        logger.setLevel(logging.DEBUG)
 
         logger.already_configured = True
     return logger

src/scgenerator/parameters.py

@@ -0,0 +1,36 @@
+class Parameter:
+    """base class for parameters"""
+
+    all = dict(fiber=dict(), pulse=dict(), gas=dict(), simulation=dict())
+    help_message = "no help message lol"
+
+    def __init_subclass__(cls, section):
+        Parameter.all[section][cls.__name__.lower()] = cls
+
+    def __init__(self, s):
+        self.s = s
+        valid = True
+        try:
+            self.value = self._convert()
+            valid = self.valid()
+        except ValueError:
+            valid = False
+
+        if not valid:
+            raise ValueError(
+                f"{self.__class__.__name__} {self.__class__.help_message}. input : {self.s}"
+            )
+
+    def _convert(self):
+        value = self.conversion_func(self.s)
+        return value
+
+
+class Wavelength(Parameter, section="pulse"):
+    help_message = "must be a strictly positive real number"
+
+    def valid(self):
+        return self.value > 0
+
+    def conversion_func(self, s: str) -> float:
+        return float(s)

src/scgenerator/physics/simulate.py

@@ -110,7 +110,7 @@ class RK4IP:
             if "raman" in self.behaviors:
                 self.conserved_quantity_func = pulse.photon_number
             else:
-                self.logger.info("energy conserved")
+                self.logger.debug("energy conserved")
                 self.conserved_quantity_func = pulse.pulse_energy
         else:
             self.conserved_quantity_func = lambda a, b, c, d: 0
@@ -179,7 +179,7 @@ class RK4IP:
     def run(self):
 
         # Print introduction
-        self.logger.info(
+        self.logger.debug(
             "Computing {} new spectra, first one at {}m".format(self.store_num, self.z_targets[0])
         )
         self.progress_tracker.set(self.z)
@@ -228,7 +228,7 @@ class RK4IP:
                 store = True
                 h_next_step = self.z_targets[0] - self.z
 
-        self.logger.info(
+        self.logger.debug(
             "propagation finished in {} steps ({} seconds)".format(
                 step, (datetime.today() - time_start).total_seconds()
             )
@@ -286,7 +286,7 @@ class RK4IP:
 
                 if curr_p_change > 2 * cons_qty_change_ok:
                     progress_str = f"step {step} rejected with h = {h:.4e}, doing over"
-                    self.logger.info(progress_str)
+                    self.logger.debug(progress_str)
                     keep = False
                     h_next_step = h / 2
                 elif cons_qty_change_ok < curr_p_change <= 2 * cons_qty_change_ok:
@@ -380,6 +380,7 @@ class Simulations:
             len(self.param_seq) * self.param_seq["simulation", "z_num"],
             percent_incr=1,
             logger=self.logger,
+            prefix="Overall : ",
         )
 
     def run(self):
@@ -478,7 +479,7 @@ class RaySimulations(Simulations, available=using_ray, priority=1):
         self.sim_jobs_per_node = min(
             self.param_seq.num_sim, self.param_seq["simulation", "parallel"]
         )
-        self.update_cluster_frequency = 5
+        self.update_cluster_frequency = 3
         self.jobs = []
         self.actors = {}
 
@@ -510,12 +511,6 @@ class RaySimulations(Simulations, available=using_ray, priority=1):
         if len(ready) == 0:
             return
         ray.get(ready)
-        # try:
-        #     ray.get(ready)
-        # except Exception as e:
-        #     self.logger.warning("A problem occured with 1 or more worker :")
-        #     self.logger.warning(e)
-        #     ray.kill(self.actors[ready[0].task_id()])
 
         del self.actors[ready[0].task_id()]
 

src/scgenerator/scripts/slurm_submit.py

@@ -0,0 +1,78 @@
+import argparse
+import os
+import re
+import shutil
+import subprocess
+from datetime import datetime, timedelta
+
+from scgenerator.initialize import validate
+from scgenerator.io import Paths, load_toml
+from scgenerator.utils import count_variations
+
+
+def format_time(t):
+    try:
+        t = float(t)
+        return timedelta(minutes=t)
+    except ValueError:
+        return t
+
+
+def create_parser():
+    parser = argparse.ArgumentParser(description="submit a job to a slurm cluster")
+    parser.add_argument("config", help="path to the toml configuration file")
+    parser.add_argument(
+        "-t", "--time", required=True, type=str, help="time required for the job in hh:mm:ss"
+    )
+    parser.add_argument(
+        "-c", "--cpus-per-node", required=True, type=int, help="number of cpus required per node"
+    )
+    parser.add_argument("-n", "--nodes", required=True, type=int, help="number of nodes required")
+    parser.add_argument(
+        "--environment-setup",
+        required=False,
+        default=f"source {os.path.expanduser('~/anaconda3/etc/profile.d/conda.sh')} && conda activate sc",
+        help="commands to run to setup the environement (default : activate the sc environment with conda)",
+    )
+    return parser
+
+
+def copy_starting_files():
+    for name in ["start_worker", "start_head"]:
+        path = Paths.get(name)
+        file_name = os.path.split(path)[1]
+        shutil.copy(path, file_name)
+        mode = os.stat(file_name)
+        os.chmod(file_name, 0o100 | mode.st_mode)
+
+
+def main():
+    parser = create_parser()
+    template = Paths.gets("submit_job_template")
+    args = parser.parse_args()
+
+    if not re.match(r"^[0-9]{2}:[0-9]{2}:[0-9]{2}$", args.time) and not re.match(
+        r"^[0-9]+$", args.time
+    ):
+
+        raise ValueError(
+            "time format must be an integer number of minute or must match the pattern hh:mm:ss"
+        )
+
+    config = load_toml(args.config)
+    config = validate(config)
+
+    sim_num, _ = count_variations(config)
+
+    file_name = "submit " + config["name"] + "-" + format(datetime.now(), "%Y%m%d%H%M") + ".sh"
+    submit_sh = template.format(**vars(args))
+    with open(file_name, "w") as file:
+        file.write(submit_sh)
+    subprocess.run(["sbatch", "--test-only", file_name])
+    submit = input(
+        f"Propagate {sim_num} pulses from config {args.config} with {args.cpus_per_node} cpus"
+        + f" per node on {args.nodes} nodes for {format_time(args.time)} ? (y/[n])\n"
+    )
+    if submit.lower() in ["y", "yes"]:
+        copy_starting_files()
+        subprocess.run(["sbatch", file_name])