Upload New File

AnalogsInnovation/ExScriptCalc.py

+import numpy as np
+import itertools
+from tqdm import tqdm
+from CalcInnovation import create_database,prediction_neighbors
+# Params
+Nreal = 10
+N = 2**16  # total number of point
+
+# Path to data
+path = '/users2/local/e22froge/codes/Julia_Innovation/DataSignals/uspa_mob15_1.dat'
+with open(path, 'rb') as fid:
+    signal = np.fromfile(fid, dtype='>f')
+data = np.reshape(signal, (-1, N))
+data = data[49:, :]
+
+# Extended execution for all parameter combinations
+overlapping_options = [False]
+random_options = [False]
+weighting_options = [True]
+normalizing_options = [True]
+p_options = [3, 5, 10, 15, 20, 30, 50, 75, 100, 200]
+Nbibpow_options = [20]
+
+for Nbibpow, p, overlapping, random, weighting, normalizing in itertools.product(
+        Nbibpow_options, p_options, overlapping_options, random_options, weighting_options, normalizing_options):
+    print(f"Running for combination: p={p}, Overlapping={overlapping}, Random={random}, Weighting={weighting}, Normalizing={normalizing}")
+
+    # Data processing and prediction for each combination
+    Bib = np.reshape(data, -1)
+    Bib = Bib[0:2**Nbibpow]
+    Bib = (Bib - np.mean(Bib)) / np.std(Bib)
+        
+    db = create_database(Bib, p, overlapping=overlapping)
+
+    vals = np.zeros((Nreal, N))
+    velocity = np.zeros((Nreal, N))
+
+    for i in tqdm(range(Nreal)):
+        real = 100 + 1 + 2 * i
+        velocity[i, :] = (data[real, :] - np.mean(data[real, :])) / np.std(data[real, :])
+        vals[i, :] = prediction_neighbors(velocity[i, :], db, k, weighting=weighting, random=random, normalize=normalizing)
+        
+    innov = velocity[:, :-1] - vals[:, :-1]
+
+    # Parameters dictionary
+    params = {
+        'Signal': 'Modane',
+        'overlapping': overlapping,
+        'random': random,
+        'weighting': weighting,
+        'normalizing': normalizing,
+        'p': p,
+        'k': k,
+        'Nreal': Nreal,
+        'N': N,
+        'Nbibpow': Nbibpow
+    }
+    # Saving the results
+    filename = f"Innov_results_Modane_N216_p{p}_k{k}_Overlap{overlapping}_Random{random}_Weight{weighting}_Norm{normalizing}_Nbibpow{Nbibpow}.npz"
+    np.savez('/users2/local/e22froge/codes/Julia_Innovation/DataInnovation/' + filename, V=velocity[:, :-1], Innov=innov, params=params)
+    print(f"Results saved to {filename}")