From f52396427043b3aceb0ccce6e752fb116e495227 Mon Sep 17 00:00:00 2001
From: Emil Karlsson <emil.karlssson@liu.se>
Date: Thu, 17 Feb 2022 21:13:23 +0100
Subject: [PATCH] Instance creation script corrected
- Incorrect file for instance file generation was previously added
---
README.md | 1 +
instances.py | 158 +++++++++++++++++++++++------------------
segmented_instances.sh | 32 ++++-----
3 files changed, 106 insertions(+), 85 deletions(-)
diff --git a/README.md b/README.md
index a9043f7..29b51b0 100644
--- a/README.md
+++ b/README.md
@@ -41,6 +41,7 @@ Instance set II is generated according to description in E. Coban and J. N. Hook
The instance generator is available in [instances.py](instances.py) and we have used Python 3.8 to generate them. To generate all instances used in our article perform the following command.
```bash
+>>> # source virtual environment with python3.8
>>> sh segmented_instances.sh
```
diff --git a/instances.py b/instances.py
index 03d76dd..84ae98a 100644
--- a/instances.py
+++ b/instances.py
@@ -2,6 +2,43 @@
A script to generate instances for the simple job shop problem with
fixed cost according to the method described in REFERENCE.
+
+alpha * R
+gamma1 * alpha * R
+gamma2 * alpha * R
+beta (dj - rj)
+
+Tight
+
+alpha = 0.5
+gamma1 = 0.5
+gamma2 = 1.0
+
+- 60: beta = 1 / 20
+- 80: beta = 1 / 20
+- 100: beta = 1 / 24
+- 120: beta = 1 / 28
+- 140: beta = 1 / 32
+- 160: beta = 1 / 32?
+- 180: beta = 1 / 32?
+- 200: beta = 1 / 32?
+
+Wide
+
+alpha = 0.5
+gamma1 = 0.25
+gamma2 = 1.0
+beta = 0.035
+
+- 60
+- 80
+- 100
+- 120
+- 140
+- 160
+- 180
+- 200
+
Usage example: python instances.py test_file.dat
"""
@@ -20,105 +57,88 @@ from collections import namedtuple
@dataclass(frozen=True, order=True)
class Job:
ID: int
- release_time: int
- deadline: int
-
+ release_time: float
+ deadline: float
+ processing_time: float
-@dataclass(frozen=True)
-class MachineJob:
- duration: int
- resource: int
- cost: int
-
-@dataclass(frozen=True)
-class Machine:
+@dataclass(frozen=True, order=True)
+class Segment:
ID: int
- resource_limit: int
-
+ start: float
+ end: float
def preparse_line(line):
return "".join(i for i in line.replace("<", "").replace("\n", "").replace(">", "").replace("[", "").replace("]", "") if i.isdigit() or i == " ")
-def generate_instance(no_jobs=None, no_machines=None, alpha=1/3, release_date_on=False):
+def generate_instance(no_jobs, alpha, gamma1, gamma2, beta, segments_percentage, time_scale):
+ TIME_SCALE = time_scale
+ segments = list()
+ assert no_jobs % round(1/segments_percentage) == 0
+ no_segments = round(segments_percentage * no_jobs) # Always ten percent of all jobs
+ R = no_segments
+ for segment_id in range(no_segments):
+ start = round(segment_id * TIME_SCALE)
+ end = round((segment_id + 1) * TIME_SCALE)
+ segment = Segment(segment_id, start, end)
+ segments.append(segment)
+
jobs = list()
- machines = list()
- machine_job_info = dict()
- capacity_limit = 10 # For all instances
- L = 5 * no_jobs * (no_machines + 1) / no_machines
for job_id in range(no_jobs):
- capacity_usage = random.randint(1, 10) # (0, 10) -> same for all machines
- max_processing_time = 0
- for machine_id in range(no_machines):
- i = machine_id + 1
- processing_time = random.randint(i, 10 * i)
- max_processing_time = max(max_processing_time, processing_time)
- cost = random.randint(2 * (no_machines - i + 1), 20 * (no_machines - i + 1))
- machine_job = MachineJob(processing_time, capacity_usage, cost)
- machine_job_info[machine_id, job_id] = machine_job
- if release_date_on:
- release_date = random.randrange(0, max_processing_time) # Same for all jobs
- deadline = release_date + max_processing_time + random.randrange(0, max_processing_time) # Same for all jobs
- else:
- release_date = 0
- deadline = alpha * L
- job = Job(job_id, release_date, deadline)
+ release_time = random.uniform(0, alpha * R)
+ deadline = release_time + random.uniform(gamma1 * alpha * R, gamma2 * alpha * R)
+ processing_time = random.uniform(0, beta * (deadline - release_time))
+ job = Job(job_id, round(TIME_SCALE * release_time), round(TIME_SCALE * deadline), round(TIME_SCALE * processing_time))
jobs.append(job)
- for machine_id in range(no_machines):
- machine = Machine(machine_id, capacity_limit)
- machines.append(machine)
- return jobs, machines, machine_job_info
+ return tuple(jobs), tuple(segments)
-def instance_to_str(jobs, machines, machine_job_info):
+def instance_to_str(jobs: Tuple[Job], segments: Tuple[Segment]):
instance_str = ""
instance_str += f"JobCount = {len(jobs)} ;\n"
- instance_str += f"MachineCount = {len(machines)} ;\n"
- instance_str += "machjob = [\n"
+ instance_str += f"SegmentCount = {len(segments)} ;\n"
+ instance_str += "Job = [\n"
- for machine in machines:
- machine_str = "["
- for job in jobs:
- machine_job = machine_job_info[machine.ID, job.ID]
- job_str = f"<{machine_job.duration} {machine_job.resource} {machine_job.cost}> "
- machine_str += job_str
- machine_str += "]\n"
- instance_str += machine_str
+ machine_str = "["
+ for job in jobs:
+ job_str = f"<{job.release_time} {job.deadline} {job.processing_time}> "
+ machine_str += job_str
+ machine_str += "]\n"
+ instance_str += machine_str
instance_str += "];\n"
- limit_str = "Limit = ["
- for machine in machines:
- limit_str += f"{machine.resource_limit} "
- limit_str += "];\n"
- instance_str += limit_str
-
- job_str = "Job = ["
- for job in jobs:
- job_str += f"<{job.release_time} {job.deadline}> "
- job_str += "]\n"
- instance_str += job_str
+ segment_str = "Segment = ["
+ for segment in segments:
+ segment_str += f"<{segment.start} {segment.end} > "
+ segment_str += "];\n"
+ instance_str += segment_str
return instance_str
+# A segment is between two consecutive integers?
if __name__ == "__main__":
# Argument parser
- parser = argparse.ArgumentParser(description='Generate an simple job shop problem with fixed cost instance')
- parser.add_argument('-r', '--release_time', action='store_true', help='Specifies if the tasks should have unique release times.')
- parser.add_argument('-m', '--no_machines', type=int, help='The number of machines', default=3)
+ parser = argparse.ArgumentParser(description='Generate a segmented feasibility problem')
+ parser.add_argument('-a', '--alpha', type=float, help='?', default=0.5)
+ parser.add_argument('-g1', '--gamma1', type=float, help='Control the lower limit of time windows', default=0.5)
+ parser.add_argument('-g2', '--gamma2', help="Control the upper limit of time windows", default=1, type=float)
parser.add_argument('-s', '--seed', help="The seed used for randomisation", default=0, type=int)
- parser.add_argument('-j', '--no_jobs', help="The number of jobs on the instance", default=20, type=int)
- parser.add_argument('-a', '--alpha', help="Selection of parameter alpha", type=float, default=0.3)
+ parser.add_argument('-j', '--no_jobs', help="The number of jobs on the instance", default=60, type=int)
+ parser.add_argument('-b', '--beta', help="Control processing time relative time windows width", type=float, default=0.035)
+ parser.add_argument('-t', '--time_scale', help="Control the time scale", type=float, default=100)
+ parser.add_argument('--segments', help="Specify the percentage of jobs that are segments", type=float, default=0.1)
parser.add_argument('instance', help='Path to instance that is created', type=str)
args = parser.parse_args()
random.seed(args.seed)
- jobs, machines, machine_job_info = generate_instance(args.no_jobs, args.no_machines, args.alpha, args.release_time)
- instance_str = instance_to_str(jobs, machines, machine_job_info)
+ jobs, segments = generate_instance(args.no_jobs, args.alpha, args.gamma1, args.gamma2, args.beta, args.segments, args.time_scale)
+ instance_str = instance_to_str(jobs, segments)
print(instance_str)
- with open(args.instance, 'w') as instance_file:
- instance_file.write(instance_str)
+ if args.instance != 'dry':
+ with open(args.instance, 'w') as instance_file:
+ instance_file.write(instance_str)
diff --git a/segmented_instances.sh b/segmented_instances.sh
index 2a2b587..bdb8871 100644
--- a/segmented_instances.sh
+++ b/segmented_instances.sh
@@ -2,25 +2,25 @@
echo "Tight windows"
for SEED in 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
do
- python instances_segmented.py "sms_segmented_instances_new/tight60j1000s$SEED.dat" --gamma1 0.0 --gamma2 0.2 --beta 0.5 --seed $SEED -j 60 -t 1000 --segments 0.10
- python instances_segmented.py "sms_segmented_instances_new/tight80j1000s$SEED.dat" --gamma1 0.0 --gamma2 0.2 --beta 0.45 --seed $SEED -j 80 -t 1000 --segments 0.10
- python instances_segmented.py "sms_segmented_instances_new/tight100j1000s$SEED.dat" --gamma1 0.0 --gamma2 0.2 --beta 0.4 --seed $SEED -j 100 -t 1000 --segments 0.10
- python instances_segmented.py "sms_segmented_instances_new/tight120j1000s$SEED.dat" --gamma1 0.0 --gamma2 0.2 --beta 0.35 --seed $SEED -j 120 -t 1000 --segments 0.10
- python instances_segmented.py "sms_segmented_instances_new/tight140j1000s$SEED.dat" --gamma1 0.0 --gamma2 0.2 --beta 0.3 --seed $SEED -j 140 -t 1000 --segments 0.10
- python instances_segmented.py "sms_segmented_instances_new/tight160j1000s$SEED.dat" --gamma1 0.0 --gamma2 0.2 --beta 0.25 --seed $SEED -j 160 -t 1000 --segments 0.10
- python instances_segmented.py "sms_segmented_instances_new/tight180j1000s$SEED.dat" --gamma1 0.0 --gamma2 0.2 --beta 0.2 --seed $SEED -j 180 -t 1000 --segments 0.10
- python instances_segmented.py "sms_segmented_instances_new/tight200j1000s$SEED.dat" --gamma1 0.0 --gamma2 0.2 --beta 0.15 --seed $SEED -j 200 -t 1000 --segments 0.10
+ python3.8 instances.py "sms_sdspmtw__instance_set_ii/tight60j1000s$SEED.dat" --gamma1 0.0 --gamma2 0.2 --beta 0.5 --seed $SEED -j 60 -t 1000 --segments 0.10
+ python3.8 instances.py "sms_sdspmtw__instance_set_ii/tight80j1000s$SEED.dat" --gamma1 0.0 --gamma2 0.2 --beta 0.45 --seed $SEED -j 80 -t 1000 --segments 0.10
+ python3.8 instances.py "sms_sdspmtw__instance_set_ii/tight100j1000s$SEED.dat" --gamma1 0.0 --gamma2 0.2 --beta 0.4 --seed $SEED -j 100 -t 1000 --segments 0.10
+ python3.8 instances.py "sms_sdspmtw__instance_set_ii/tight120j1000s$SEED.dat" --gamma1 0.0 --gamma2 0.2 --beta 0.35 --seed $SEED -j 120 -t 1000 --segments 0.10
+ python3.8 instances.py "sms_sdspmtw__instance_set_ii/tight140j1000s$SEED.dat" --gamma1 0.0 --gamma2 0.2 --beta 0.3 --seed $SEED -j 140 -t 1000 --segments 0.10
+ python3.8 instances.py "sms_sdspmtw__instance_set_ii/tight160j1000s$SEED.dat" --gamma1 0.0 --gamma2 0.2 --beta 0.25 --seed $SEED -j 160 -t 1000 --segments 0.10
+ python3.8 instances.py "sms_sdspmtw__instance_set_ii/tight180j1000s$SEED.dat" --gamma1 0.0 --gamma2 0.2 --beta 0.2 --seed $SEED -j 180 -t 1000 --segments 0.10
+ python3.8 instances.py "sms_sdspmtw__instance_set_ii/tight200j1000s$SEED.dat" --gamma1 0.0 --gamma2 0.2 --beta 0.15 --seed $SEED -j 200 -t 1000 --segments 0.10
done
echo "Wide windows"
for SEED in 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
do
- python instances_segmented.py "sms_segmented_instances_new/wide60j1000s$SEED.dat" --gamma1 0.0 --gamma2 0.5 --beta 0.5 --seed $SEED -j 60 -t 1000 --segments 0.10
- python instances_segmented.py "sms_segmented_instances_new/wide80j1000s$SEED.dat" --gamma1 0.0 --gamma2 0.5 --beta 0.45 --seed $SEED -j 80 -t 1000 --segments 0.10
- python instances_segmented.py "sms_segmented_instances_new/wide100j1000s$SEED.dat" --gamma1 0.0 --gamma2 0.5 --beta 0.4 --seed $SEED -j 100 -t 1000 --segments 0.10
- python instances_segmented.py "sms_segmented_instances_new/wide120j1000s$SEED.dat" --gamma1 0.0 --gamma2 0.5 --beta 0.35 --seed $SEED -j 120 -t 1000 --segments 0.10
- python instances_segmented.py "sms_segmented_instances_new/wide140j1000s$SEED.dat" --gamma1 0.0 --gamma2 0.5 --beta 0.30 --seed $SEED -j 140 -t 1000 --segments 0.10
- python instances_segmented.py "sms_segmented_instances_new/wide160j1000s$SEED.dat" --gamma1 0.0 --gamma2 0.5 --beta 0.25 --seed $SEED -j 160 -t 1000 --segments 0.10
- python instances_segmented.py "sms_segmented_instances_new/wide180j1000s$SEED.dat" --gamma1 0.0 --gamma2 0.5 --beta 0.20 --seed $SEED -j 180 -t 1000 --segments 0.10
- python instances_segmented.py "sms_segmented_instances_new/wide200j1000s$SEED.dat" --gamma1 0.0 --gamma2 0.5 --beta 0.15 --seed $SEED -j 200 -t 1000 --segments 0.10
+ python3.8 instances.py "sms_sdspmtw__instance_set_ii/wide60j1000s$SEED.dat" --gamma1 0.0 --gamma2 0.5 --beta 0.5 --seed $SEED -j 60 -t 1000 --segments 0.10
+ python3.8 instances.py "sms_sdspmtw__instance_set_ii/wide80j1000s$SEED.dat" --gamma1 0.0 --gamma2 0.5 --beta 0.45 --seed $SEED -j 80 -t 1000 --segments 0.10
+ python3.8 instances.py "sms_sdspmtw__instance_set_ii/wide100j1000s$SEED.dat" --gamma1 0.0 --gamma2 0.5 --beta 0.4 --seed $SEED -j 100 -t 1000 --segments 0.10
+ python3.8 instances.py "sms_sdspmtw__instance_set_ii/wide120j1000s$SEED.dat" --gamma1 0.0 --gamma2 0.5 --beta 0.35 --seed $SEED -j 120 -t 1000 --segments 0.10
+ python3.8 instances.py "sms_sdspmtw__instance_set_ii/wide140j1000s$SEED.dat" --gamma1 0.0 --gamma2 0.5 --beta 0.30 --seed $SEED -j 140 -t 1000 --segments 0.10
+ python3.8 instances.py "sms_sdspmtw__instance_set_ii/wide160j1000s$SEED.dat" --gamma1 0.0 --gamma2 0.5 --beta 0.25 --seed $SEED -j 160 -t 1000 --segments 0.10
+ python3.8 instances.py "sms_sdspmtw__instance_set_ii/wide180j1000s$SEED.dat" --gamma1 0.0 --gamma2 0.5 --beta 0.20 --seed $SEED -j 180 -t 1000 --segments 0.10
+ python3.8 instances.py "sms_sdspmtw__instance_set_ii/wide200j1000s$SEED.dat" --gamma1 0.0 --gamma2 0.5 --beta 0.15 --seed $SEED -j 200 -t 1000 --segments 0.10
done
--
GitLab