Skip to content

Latest commit

 

History

History
30 lines (21 loc) · 2.65 KB

README.md

File metadata and controls

30 lines (21 loc) · 2.65 KB
Raw

The .py files in this folder are the parameter files that can be used to load the pre-tuned parameters to re-produce the results shown in the paper.

The train_and_test.py loads the required parameters to run, load and save the network and results from the params.py. The params.py file shows an example implemetation of these parameters with description of each parameter's use in the network. This folder contains the pre-tuned parameter files for re-producing the results shown in the paper.

Description of the parameter files

Here is a description of what parameters were used for each result presented in the paper.

Please refer to DLoc paper to see the figure numbers and the appropriate results that you can reproduce

  • params_fig10a.py: Generate the DLoc results to reproduce the plot in Figure 10a
  • params_fig10b.py: Generate the DLoc results to reproduce the plot in Figure 10b
  • params_fig11a.py: Generate the DLoc results to reproduce the plot in Figure 11a for without copensation decoder
  • params_fig11b.py: Generate the DLoc results to reproduce the plot in Figure 11b for without copensation decoder
  • params_fig13a_20MHz.py: Generate the DLoc results to reproduce the plot in Figure 13a for the 20MHz Bandwidth
  • params_fig13a_40MHz.py: Generate the DLoc results to reproduce the plot in Figure 13a for the 40MHz Bandwidth
  • params_fig13b.py: Generate the DLoc results to reproduce the plot in Figure 13b for the Disjoint dataset
  • params_tab1_test2.py: Generate the DLoc results to reproduce the results in Table1, where the network is trained on Env-1/3/4 and Tested on Env-2
  • params_tab1_test3.py: Generate the DLoc results to reproduce the results in Table1, where the network is trained on Env-1/2/4 and Tested on Env-3
  • params_tab1_test4.py: Generate the DLoc results to reproduce the results in Table1, where the network is trained on Env-1/2/3 and Tested on Env-4

Loading the parameters

As mentioned earlier, train_and_test.py file loads the required parameters from params.py. So to load the pre-tuned parameters to re-generate the results from the paper, please copy the correponding parameter files in this folder to params.py file in the parent folder.

cp <params_file.py> ../params.py

Then you can run the train_and_test.py script to re-produce the results from the paper.