sched_group_ws_sample

Package for RD-Gen and SchedulingVisualizer, and Templete for program of thesis. Sample procedures are described in each jupyter notebook (.ipynb).

0. install dependency package

pip3 install -r ./requirements.txt

1. RD-Gen Reader

get DAG status from RD-Gen (https://github.com/azu-lab/RD-Gen)

RDG_DAG takes as input the array of execution times and the array of edges

in addition, methods related to graph structures use the networkx library

for detail, please refer src/dag_rd_gen.py

1. 1. Basic status of DAG

1. 2. Graph status of DAG

2. Utility of Scheduling_Visualization

visualize scheduling result

visualisation is done in an external library (https://github.com/atsushi421/Scheduling_Visualization)

two methods have been implemented: to prepare objects for visualisation and to provide arrays of data

2. 1. Generate visualizing target

2. 2. visualize from array

3. RD-Gen -> scheduling -> Scheduling_Visualization

minimum workflow of scheduling simulation

scheduler/ directory is a simple implementation of list scheduling

For details of sample list scheduling, refer to the following: https://github.com/rokamu623/list_scheduling

4. evaluate taskset

a sample of evaluate scheduling algorithm

implementation is refered in src/simulate.py

4. 1. load RD-Gen from directory

4. 2. evaluate taskset

scheduling algorithm is implemented in src.simulate.simulate function

src.simulate.simulate return evaluated value (ex. WCRT)

src.simulate.simulate_taskset return list of evaluated values for all DAG (dags)

4. 2'. evaluate taskset

Instead of using the simulate_taskset function, the unit tests are written in .py files and all loops are done in .ipynb

4. 3. visualize

↓ appendix (not refactored) ↓

5. prepare to existing method

5. 1. load RD-Gen from directory

5. 2. prepare to existing method

6. evaluate with varing parameter (extra)

6. 1. load RD-Gen from directory

6. 2. evaluate with varing parameter

6. 3. visualize

directory

• output/ : graphs and files are outputted here
• sample_input/ : input files of samples are here
• sample_shceduler/ : scheduling algorithm using sample (not refactored)
• src/
  • dag_rd_gen.py : RDG_DAG class which can load data from RD-Gen file is implemented
  • sched_vis_util.py: Visualize helper class is implemented
  • simulate.py : sample function of evaluate scheduling algorithm
• n.xxx.ipynb : sample of function of this repository
• README.md : this
• requirements.txt : package list of requirement packages