| import numpy as np |
| import matplotlib.pyplot as plt |
| import matplotlib |
| import os |
| from itertools import chain, combinations |
|
|
| matplotlib.rcParams['pdf.fonttype'] = 42 |
| matplotlib.rcParams['ps.fonttype'] = 42 |
|
|
|
|
| def powerset(iterable): |
| "powerset([1,2,3]) --> () (1,) (2,) (3,) (1,2) (1,3) (2,3) (1,2,3)" |
| s = list(iterable) |
| return chain.from_iterable(combinations(s, r) for r in range(len(s) + 1)) |
|
|
|
|
| |
| base_path = "./workspace/Jammer_detection/Src/federated_synthetic_data/federated_scr/FedAvgM_53mmd/model/" |
| output_folder = os.path.join(base_path, "Test_round_10") |
| summary_path = base_path + "summary_all_round.txt" |
| if not os.path.exists(output_folder): |
| os.makedirs(output_folder) |
| with open(summary_path, "r") as f: |
| data_raw = f.readlines() |
|
|
| results = [] |
| methods = [] |
| swap = 0 |
| cur_results = [] |
| cur_results_comb = [] |
| for di, dr in enumerate(data_raw): |
| dr_split = dr.split("\n")[0].split("\t") |
| if len(dr_split) == 1: |
| methods.append(dr_split[0]) |
| swap = -1 |
| if di != 0: |
| results.append(cur_results_comb) |
| cur_results_comb = [] |
| else: |
| cur_results.append([float(x) for x in dr_split[1:-1] if len(x) != 0]) |
| if swap == 1: |
| cur_results_comb.append(cur_results) |
| cur_results = [] |
| swap = (swap + 1) % 2 |
| results.append(cur_results_comb) |
| results = np.array(results) |
| print(f"shape od result is ----------{results.shape}") |
| additional_classes = [3 + np.array(list(x)).astype(int) for x in |
| list(powerset("".join((np.array([3, 4, 5]) - 3).astype("str"))))] |
| print(additional_classes) |
| font = {'family': 'serif', |
| 'weight': 'bold', |
| 'size': 10} |
| for i, res in enumerate(results): |
| trained_on = np.array([0, 1, 2]) |
| trained_on.sort() |
| img = np.around(res[:, 0, 2].reshape((4, 2)) * 100, 1) |
| img_arg = np.around(res[:, 1, 2].reshape((4, 2)) * 100, 1) |
| prec_rec = np.around(res[:, 0, 3:].reshape((4, 2, 2)) * 100, 1).astype("str") |
| plt.figure(figsize=(10, 5)) |
| plt.title("Trained on: " + methods[i].split("_")[0]) |
| plt.imshow(img, cmap="rainbow", vmin=0, vmax=100) |
| for ci in range(8): |
| cii = ci // 2 |
| cij = ci % 2 |
| |
| ac = " ".join(additional_classes[ci].astype("str")) + "\n" + str(img[cii, cij]) + "\n" + " ".join( |
| prec_rec[cii, cij]) |
| plt.text(cij, cii, ac, ha='center', va='center', **font) |
| print(np.arange(-0.5, 4.0, 1.0)) |
| print(np.arange(5)) |
| plt.xticks(np.arange(-0.5, 4.0, 1.0)[:2], "") |
| plt.yticks(np.arange(-0.5, 4.0, 1.0), "") |
| plt.colorbar() |
| plt.grid(color='k', linestyle='-', linewidth=2) |
| plt.tight_layout() |
| plt.savefig(os.path.join(output_folder, "Res%d.png" % i)) |
| |
| plt.close() |
|
|
