Feature periodic score

README.md

@@ -271,7 +271,8 @@ Several flags can be specified. The most relevant settings are:
 | -d, --dtw                                           | performs dynamic time warping on the top 3 frequencies (highest contribution) calculated using the DFT if set (default=False)                                                                                                                                                                                                                                                                                             |
 | -re, --reconstruction                               | plots reconstruction of top 10 signals on figure                                                                                                                                                                                                                                                                                                                                                                          |
 | -np, --no-psd                                       | if set, replace the power density spectrum (a*a/N) with the amplitude spectrum (a)                                                                                                                                                                                                                                                                                                                                        |
-| -c, --autocorrelation                               | if set, autocorrelation is calculated in addition to DFT. The results are merged to a single prediction at the end                                                                                                                                                                                                                                                                                                        |
+| -au, --autocorrelation                              | if set, autocorrelation is calculated in addition to DFT. The results are merged to a single prediction at the end                                                                                                                                                                                                                                                                                                        |
+| -p, --periodicity                                   | Activate calculation of new periodicity score. Options are recurrence period density entropy (RPDE), spectral flatness (SF), correlation (corr) and individual period correlation (ind)                                                                                                                                                                                                                                   |
 | -w, --window_adaptation                             | online time window adaptation. If set to true, the time window is shifted on X hits to X times the previous phases from the current instance. X corresponds to frequency_hits                                                                                                                                                                                                                                             |
 | -fh FREQUENCY_HITS, --frequency_hits FREQUENCY_HITS | specifies the number of hits needed to adapt the time window. A hit occurs once a dominant frequency is found                                                                                                                                                                                                                                                                                                             |
 | -v, --verbose                                       | sets verbose on or off (default=False)                                                                                                                                                                                                                                                                                                                                                                                    |
@@ -391,4 +392,4 @@ the [EuroHPC ADMIRE project](https://admire-eurohpc.eu/).
 
 [parallel.bedge]: https://img.shields.io/badge/Parallel_Programming:-Ahmad_Tarraf-blue
 
-<!--* links *-->
+<!--* links *-->

daf

@@ -0,0 +1,32 @@
+commit bc2c874a2166cbc0ab6740e83dbd6d1ea8f2d28c (HEAD -> feature_periodic_score, origin/feature_periodic_score)
+Merge: f176334 34a108b
+Author: AntonBeasis <anton@holderied.de>
+Date:   Tue Jun 3 11:13:19 2025 +0200
+
+    Merge completed
+
+ create mode 100644 docs/contributing.md
+ create mode 100644 docs/ftio_server.md
+ create mode 100644 ftio/analysis/__init__.py
+ create mode 100644 ftio/analysis/_correlation.py
+ rename ftio/{freq => analysis}/_logicize.py (55%)
+ rename ftio/{freq => analysis}/anomaly_detection.py (80%)
+ create mode 100644 ftio/analysis/signal_analysis.py
+ create mode 100644 ftio/api/gekkoFs/data_control.py
+ create mode 100644 ftio/api/gekkoFs/file_queue.py
+ create mode 100644 ftio/api/gekkoFs/jit/find_procs.py
+ create mode 100644 ftio/api/gekkoFs/jit/plot_res.py
+ create mode 100644 ftio/api/gekkoFs/posix_control.py
+ create mode 100644 ftio/api/io_malleability/diff.py
+ create mode 100644 ftio/freq/_analysis_figures.py
+ create mode 100644 ftio/freq/_dft_x_dwt.py
+ create mode 100644 ftio/freq/_fourier_fit.py
+ create mode 100644 ftio/freq/_share_signal_data.py
+ create mode 100644 ftio/freq/prediction.py
+ create mode 100644 ftio/plot/plot_dft.py
+ delete mode 100644 ftio/prediction/analysis.py
+ create mode 100644 ftio/prediction/online_analysis.py
+ rename ftio/processing/{operations.py => compact_operations.py} (74%)
+ create mode 100644 ftio/processing/print_output.py
+ create mode 100644 ftio/util/live_plot.py
+ create mode 100644 ftio/util/server_ftio.py

docs/contributing.md

@@ -90,4 +90,5 @@ By contributing, you agree that your contributions will be licensed under the sa
 
 We sincerely thank the following contributors for their valuable contributions:
 - [Ahmad Tarraf](https://github.com/a-tarraf)
-- [Jean-Baptiste Bensard](https://github.com/besnardjb): Metric proxy integration
+- [Jean-Baptiste Bensard](https://github.com/besnardjb): Metric proxy integration
+- [Anton Holderied](https://github.com/AntonBeasis): bachelor thesis: new periodicity score

examples/jupyter_notebooks/quick_example.ipynb

@@ -18,10 +18,55 @@
     "\n",
     "# Set up data\n",
     "## 1) overlap for rank level metrics\n",
-    "b_rank   = [0.0,0.0,1000.0,1000.0,0.0,0.0,1000.0,1000.0,0.0,0.0,1000.0,1000.0,0.0,0.0]\n",
-    "t_rank_s = [0.5,0.0,10.5,10.0,20.5,20.0,30.5,30.0,40.5,40.0,50.5,50.0,60.5,60]\n",
-    "t_rank_e = [5.0,4.5,15.0,14.5,25.0,24.5,35.0,34.5,45.0,44.5,55.0,54.5,65.0,64.5]\n",
-    "b,t = overlap(b_rank,t_rank_s, t_rank_e)\n",
+    "b_rank = [\n",
+    "    0.0,\n",
+    "    0.0,\n",
+    "    1000.0,\n",
+    "    1000.0,\n",
+    "    0.0,\n",
+    "    0.0,\n",
+    "    1000.0,\n",
+    "    1000.0,\n",
+    "    0.0,\n",
+    "    0.0,\n",
+    "    1000.0,\n",
+    "    1000.0,\n",
+    "    0.0,\n",
+    "    0.0,\n",
+    "]\n",
+    "t_rank_s = [\n",
+    "    0.5,\n",
+    "    0.0,\n",
+    "    10.5,\n",
+    "    10.0,\n",
+    "    20.5,\n",
+    "    20.0,\n",
+    "    30.5,\n",
+    "    30.0,\n",
+    "    40.5,\n",
+    "    40.0,\n",
+    "    50.5,\n",
+    "    50.0,\n",
+    "    60.5,\n",
+    "    60,\n",
+    "]\n",
+    "t_rank_e = [\n",
+    "    5.0,\n",
+    "    4.5,\n",
+    "    15.0,\n",
+    "    14.5,\n",
+    "    25.0,\n",
+    "    24.5,\n",
+    "    35.0,\n",
+    "    34.5,\n",
+    "    45.0,\n",
+    "    44.5,\n",
+    "    55.0,\n",
+    "    54.5,\n",
+    "    65.0,\n",
+    "    64.5,\n",
+    "]\n",
+    "b, t = overlap(b_rank, t_rank_s, t_rank_e)\n",
     "\n",
     "# ## 2) or directly specify the app level metrics\n",
     "# t = [10.0, 20.1, 30.0, 40.2, 50.3, 60, 70, 80.0,]\n",
@@ -33,19 +78,19 @@
     "\n",
     "# set up data\n",
     "data = {\n",
-    "        \"time\": np.array(t),\n",
-    "        \"bandwidth\": np.array(b),\n",
-    "        \"total_bytes\": total_bytes,\n",
-    "        \"ranks\": ranks \n",
-    "        }\n",
+    "    \"time\": np.array(t),\n",
+    "    \"bandwidth\": np.array(b),\n",
+    "    \"total_bytes\": total_bytes,\n",
+    "    \"ranks\": ranks,\n",
+    "}\n",
     "\n",
-    "#parse args\n",
-    "args = parse_args(argv,\"ftio\")\n",
+    "# parse args\n",
+    "args = parse_args(argv, \"ftio\")\n",
     "\n",
     "# perform prediction\n",
     "prediction, analysis_figures = core(data, args)\n",
     "\n",
-    "#display prediction\n",
+    "# display prediction\n",
     "display_prediction(\"ftio\", prediction)"
    ]
   },
@@ -76,29 +121,74 @@
     "\n",
     "# Set up data\n",
     "## 1) overlap for rank level metrics\n",
-    "b_rank   = [0.0,0.0,1000.0,1000.0,0.0,0.0,1000.0,1000.0,0.0,0.0,1000.0,1000.0,0.0,0.0]\n",
-    "t_rank_s = [0.5,0.0,10.5,10.0,20.5,20.0,30.5,30.0,40.5,40.0,50.5,50.0,60.5,60]\n",
-    "t_rank_e = [5.0,4.5,15.0,14.5,25.0,24.5,35.0,34.5,45.0,44.5,55.0,54.5,65.0,64.5]\n",
-    "b,t = overlap(b_rank,t_rank_s, t_rank_e)\n",
+    "b_rank = [\n",
+    "    0.0,\n",
+    "    0.0,\n",
+    "    1000.0,\n",
+    "    1000.0,\n",
+    "    0.0,\n",
+    "    0.0,\n",
+    "    1000.0,\n",
+    "    1000.0,\n",
+    "    0.0,\n",
+    "    0.0,\n",
+    "    1000.0,\n",
+    "    1000.0,\n",
+    "    0.0,\n",
+    "    0.0,\n",
+    "]\n",
+    "t_rank_s = [\n",
+    "    0.5,\n",
+    "    0.0,\n",
+    "    10.5,\n",
+    "    10.0,\n",
+    "    20.5,\n",
+    "    20.0,\n",
+    "    30.5,\n",
+    "    30.0,\n",
+    "    40.5,\n",
+    "    40.0,\n",
+    "    50.5,\n",
+    "    50.0,\n",
+    "    60.5,\n",
+    "    60,\n",
+    "]\n",
+    "t_rank_e = [\n",
+    "    5.0,\n",
+    "    4.5,\n",
+    "    15.0,\n",
+    "    14.5,\n",
+    "    25.0,\n",
+    "    24.5,\n",
+    "    35.0,\n",
+    "    34.5,\n",
+    "    45.0,\n",
+    "    44.5,\n",
+    "    55.0,\n",
+    "    54.5,\n",
+    "    65.0,\n",
+    "    64.5,\n",
+    "]\n",
+    "b, t = overlap(b_rank, t_rank_s, t_rank_e)\n",
     "\n",
     "# command line arguments\n",
     "argv = [\"-e\", \"plotly\"]\n",
     "\n",
     "# set up data\n",
     "data = {\n",
-    "        \"time\": np.array(t),\n",
-    "        \"bandwidth\": np.array(b),\n",
-    "        \"total_bytes\": total_bytes,\n",
-    "        \"ranks\": ranks \n",
-    "        }\n",
+    "    \"time\": np.array(t),\n",
+    "    \"bandwidth\": np.array(b),\n",
+    "    \"total_bytes\": total_bytes,\n",
+    "    \"ranks\": ranks,\n",
+    "}\n",
     "\n",
-    "#parse args\n",
-    "args = parse_args(argv,\"ftio\")\n",
+    "# parse args\n",
+    "args = parse_args(argv, \"ftio\")\n",
     "\n",
     "# perform prediction\n",
     "prediction, analysis_figures = core(data, args)\n",
     "\n",
-    "#display prediction\n",
+    "# display prediction\n",
     "display_prediction(\"ftio\", prediction)\n",
     "analysis_figures.show()"
    ]

ftio/analysis/anomaly_detection.py

@@ -30,6 +30,7 @@
 from sklearn.neighbors import LocalOutlierFactor, NearestNeighbors
 
 from ftio.plot.anomaly_plot import plot_decision_boundaries, plot_outliers
+from ftio.plot.cepstrum_plot import plot_cepstrum
 
 
 def outlier_detection(
@@ -164,6 +165,8 @@ def z_score(
         if len(dominant_index) != 0:
             i[np.array(dominant_index) - 1] = -1
         plot_outliers(args, freq_arr, amp, indices, conf, i)
+        if args.cepstrum:
+            plot_cepstrum(args, freq_arr, amp, indices)
 
     return dominant_index, conf, text