change the way the normalized distance is calculated

d216708b · Maël Madon · d07005f3 · d216708b · d216708b
Commit d216708b authored 2 years ago by Maël Madon
--- a/distance_batsim_output.py
+++ b/distance_batsim_output.py
@@ -49,8 +49,20 @@ def open_and_compare(file1, file2):
    out1 = out1[out1.success == 1]
    out2 = out2[out2.success == 1]

+    runtimes1 = out1.finish_time - out1.starting_time
+    runtimes2 = out2.finish_time - out2.starting_time
+    
+    if not np.allclose(runtimes1, runtimes2, atol=1):
+        warnings.warn(f"Some jobs in {file1} and {file2} don't have the same runtime (+/- 1sec). It is unusual, as runtime are normally an input of the simulation.")
+
    return out1, out2

+def sum_execution_times(out):
+    """Returns the sum of all job execution times from a clean batsim output"""
+
+    runtimes = out.finish_time - out.starting_time
+    return runtimes.sum()
+

 def euclidean_distance(s1, s2):
    """Returns the Euclidean distance between two series s1 and s2"""
@@ -64,15 +76,6 @@ def lateness_distance(s1, s2):

    return float(np.sum([y-x for x, y in zip(s1, s2)]))

-def normalized_euclidian_distance(s1, s2):
-    """Return the euclidien distance normalized by the l2 norm of the vectors, 
-    or None if one of the vectors is the null vector (undefined)"""
-
-    n1, n2 = l2_norm(s1), l2_norm(s2)
-    if n1==0 or n2==0:
-        return None
-    eucl_dist = euclidean_distance(s1, s2)
-    return float( eucl_dist**2 / (n1 * n2) )

 def l2_norm(s):
    """Return the l2 norm of the series s"""
@@ -87,13 +90,20 @@ def distances(file1, file2, euclidean=True, lateness=False, norm_eucl=False,

    out1, out2 = open_and_compare(file1, file2)

+    sum_runtimes1 = sum_execution_times(out1)
+    sum_runtimes2 = sum_execution_times(out2)
+
    dist = {}
    for f in field:
        dist[f] = {}
+        eucl = euclidean_distance(out1[f], out2[f])
        if euclidean:
-            dist[f]["euclidean"] = euclidean_distance(out1[f], out2[f])
+            dist[f]["euclidean"] = eucl
        if norm_eucl:
-            dist[f]["normalized_euclidean"] = normalized_euclidian_distance(out1[f], out2[f])
+            if sum_runtimes1 == 0 or sum_runtimes2 == 0:
+                dist[f]["normalized_euclidean"] = None
+            else:
+                dist[f]["normalized_euclidean"] = eucl / np.sqrt(sum_runtimes1 * sum_runtimes2)
        if lateness:
            dist[f]["lateness"] = lateness_distance(out1[f], out2[f])


--- a/example_distance.ipynb
+++ b/example_distance.ipynb
@@ -37,7 +37,7 @@
  },
  {
   "cell_type": "code",
-   "execution_count": 3,
+   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
@@ -179,7 +179,7 @@
       "[324 rows x 5 columns]"
      ]
     },
-     "execution_count": 3,
+     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
@@ -250,13 +250,15 @@
  },
  {
   "cell_type": "code",
-   "execution_count": 6,
+   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
+      "/home/mael/ownCloud/workspace/batmen-tools/distance_batsim_output.py:56: UserWarning: Some jobs in test/input/3jobs.csv and test/input/3jobs_w_sessions.csv don't have the same runtime (+/- 1sec). It is unusual, as runtime are normally an input of the simulation.\n",
+      "  warnings.warn(f\"Some jobs in {file1} and {file2} don't have the same runtime (+/- 1sec). It is unusual, as runtime are normally an input of the simulation.\")\n",
      "{\n",
      "    \"submission_time\": {\n",
      "        \"euclidean\": 0.0,\n",
@@ -270,7 +272,7 @@
      "    },\n",
      "    \"finish_time\": {\n",
      "        \"euclidean\": 28.284271247461902,\n",
-      "        \"normalized_euclidean\": 0.08000000000000002,\n",
+      "        \"normalized_euclidean\": 0.4040610178208843,\n",
      "        \"lateness\": 0.0\n",
      "    }\n",
      "}\n"
@@ -283,27 +285,31 @@
  },
  {
   "cell_type": "code",
-   "execution_count": 7,
+   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
+      "/home/mael/ownCloud/workspace/batmen-tools/distance_batsim_output.py:56: UserWarning: Some jobs in test/input/3jobs.csv and test/input/3jobs_zeros.csv don't have the same runtime (+/- 1sec). It is unusual, as runtime are normally an input of the simulation.\n",
+      "  warnings.warn(f\"Some jobs in {file1} and {file2} don't have the same runtime (+/- 1sec). It is unusual, as runtime are normally an input of the simulation.\")\n",
+      "/home/mael/ownCloud/workspace/batmen-tools/distance_batsim_output.py:114: RuntimeWarning: divide by zero encountered in double_scalars\n",
+      "  dist[f][\"normalized_euclidean\"] = eucl / np.sqrt(sum_runtimes1 * sum_runtimes2)\n",
      "{\n",
      "    \"submission_time\": {\n",
      "        \"euclidean\": 50.0,\n",
-      "        \"normalized_euclidean\": null,\n",
+      "        \"normalized_euclidean\": Infinity,\n",
      "        \"lateness\": -70.0\n",
      "    },\n",
      "    \"starting_time\": {\n",
      "        \"euclidean\": 50.0,\n",
-      "        \"normalized_euclidean\": null,\n",
+      "        \"normalized_euclidean\": Infinity,\n",
      "        \"lateness\": -70.0\n",
      "    },\n",
      "    \"finish_time\": {\n",
      "        \"euclidean\": 100.0,\n",
-      "        \"normalized_euclidean\": null,\n",
+      "        \"normalized_euclidean\": Infinity,\n",
      "        \"lateness\": -140.0\n",
      "    }\n",
      "}\n"
@@ -316,7 +322,7 @@
  },
  {
   "cell_type": "code",
-   "execution_count": 8,
+   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
@@ -326,17 +332,17 @@
      "{\n",
      "    \"submission_time\": {\n",
      "        \"euclidean\": 172.1394783308001,\n",
-      "        \"normalized_euclidean\": 2.739806304756353e-10,\n",
+      "        \"normalized_euclidean\": 1.2942984329407551e-05,\n",
      "        \"lateness\": 2254.0\n",
      "    },\n",
      "    \"starting_time\": {\n",
      "        \"euclidean\": 172.1394783308001,\n",
-      "        \"normalized_euclidean\": 2.739806304756353e-10,\n",
+      "        \"normalized_euclidean\": 1.2942984329407551e-05,\n",
      "        \"lateness\": 2254.0\n",
      "    },\n",
      "    \"finish_time\": {\n",
      "        \"euclidean\": 172.1394783308001,\n",
-      "        \"normalized_euclidean\": 2.3852456580198333e-10,\n",
+      "        \"normalized_euclidean\": 1.2942984329407551e-05,\n",
      "        \"lateness\": 2254.0\n",
      "    }\n",
      "}\n"
@@ -349,7 +355,7 @@
  },
  {
   "cell_type": "code",
-   "execution_count": 9,
+   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
@@ -359,17 +365,17 @@
      "{\n",
      "    \"submission_time\": {\n",
      "        \"euclidean\": 241.15762480170517,\n",
-      "        \"normalized_euclidean\": 5.377222025784075e-10,\n",
+      "        \"normalized_euclidean\": 1.813216587526566e-05,\n",
      "        \"lateness\": 3311.0\n",
      "    },\n",
      "    \"starting_time\": {\n",
      "        \"euclidean\": 241.15762480170517,\n",
-      "        \"normalized_euclidean\": 5.377222025784075e-10,\n",
+      "        \"normalized_euclidean\": 1.813216587526566e-05,\n",
      "        \"lateness\": 3311.0\n",
      "    },\n",
      "    \"finish_time\": {\n",
      "        \"euclidean\": 255.14897609044016,\n",
-      "        \"normalized_euclidean\": 5.240304046302593e-10,\n",
+      "        \"normalized_euclidean\": 1.91841479662116e-05,\n",
      "        \"lateness\": 3633.0\n",
      "    }\n",
      "}\n"

 %% Cell type:markdown id: tags:

 # Example of distances

 This notebook calculate distances between several _jobs.csv as an example.

 %% Cell type:code id: tags:

 ``` python
 # Initialization
 import pandas as pd
 from evalys.jobset import JobSet
 from evalys.visu.gantt import plot_gantt

 three_jobs =            "test/input/3jobs.csv"
 three_jobs_w_session =  "test/input/3jobs_w_sessions.csv"
 three_jobs_zero =       "test/input/3jobs_zeros.csv"
 mc_10days_a60 =         "test/input/mc_10days_a60_jobs.csv"
 mc_10days_m60 =         "test/input/mc_10days_m60_jobs.csv"
 mc_10days_rigid =       "test/input/mc_10days_rigid_jobs.csv"
 ```

 %% Cell type:markdown id: tags:

 Vizualize the useful columns of a jobs.csv:

 %% Cell type:code id: tags:

 ``` python
 out1 = pd.read_csv(mc_10days_a60)

 # Select
 desired_cols = ["job_id", "submission_time", "starting_time", "finish_time", "success"]
 select = out1.loc[:, desired_cols]

 # Clean job_id
 select.job_id = select.job_id.astype(str)
 select.job_id = select.job_id.str.split(':', expand=True)[0]
 select.job_id = select.job_id.astype(int)

 select
 ```

 %% Output

         job_id  submission_time  starting_time  finish_time  success
    0      1216          55532.0        55532.0      55894.0        1
    1       247           9327.0         9327.0      66429.0        1
    2      1242          56876.0        56876.0      66496.0        1
    3      1434          66504.0        66504.0      67496.0        1
    4      1438          66506.0        66506.0      69764.0        1
    ..      ...              ...            ...          ...      ...
    319   18943         643942.0       643942.0     954600.0        1
    320   18945         643943.0       643943.0     958295.0        1
    321   21862         753752.0       753752.0     960982.0        1
    322   18944         643942.0       643942.0     961174.0        1
    323   21995         760446.0       760446.0    1031099.0        1
    
    [324 rows x 5 columns]

 %% Cell type:markdown id: tags:

 ## With mock files
 Visualize our two mock files 3jobs and 3jobs_w_sessions:

 %% Cell type:code id: tags:

 ``` python
 plot_gantt(JobSet.from_csv(three_jobs), title="three_jobs")
 plot_gantt(JobSet.from_csv(three_jobs_w_session), title="three_jobs_w_session")
 ```

 %% Output

    /home/mael/.local/lib/python3.10/site-packages/evalys/visu/core.py:62: UserWarning: Matplotlib is currently using module://matplotlib_inline.backend_inline, which is a non-GUI backend, so cannot show the figure.
      self.fig.show()
    /home/mael/.local/lib/python3.10/site-packages/evalys/visu/core.py:62: UserWarning: Matplotlib is currently using module://matplotlib_inline.backend_inline, which is a non-GUI backend, so cannot show the figure.
      self.fig.show()





 %% Cell type:code id: tags:

 ``` python
 !python3 distance_batsim_output.py {three_jobs} {three_jobs_w_session} --all
 ```

 %% Output

+    /home/mael/ownCloud/workspace/batmen-tools/distance_batsim_output.py:56: UserWarning: Some jobs in test/input/3jobs.csv and test/input/3jobs_w_sessions.csv don't have the same runtime (+/- 1sec). It is unusual, as runtime are normally an input of the simulation.
+      warnings.warn(f"Some jobs in {file1} and {file2} don't have the same runtime (+/- 1sec). It is unusual, as runtime are normally an input of the simulation.")
    {
        "submission_time": {
            "euclidean": 0.0,
            "normalized_euclidean": 0.0,
            "lateness": 0.0
        },
        "starting_time": {
            "euclidean": 0.0,
            "normalized_euclidean": 0.0,
            "lateness": 0.0
        },
        "finish_time": {
            "euclidean": 28.284271247461902,
-            "normalized_euclidean": 0.08000000000000002,
+            "normalized_euclidean": 0.4040610178208843,
            "lateness": 0.0
        }
    }

 %% Cell type:code id: tags:

 ``` python
 !python3 distance_batsim_output.py {three_jobs} {three_jobs_zero} --all
 ```

 %% Output

+    /home/mael/ownCloud/workspace/batmen-tools/distance_batsim_output.py:56: UserWarning: Some jobs in test/input/3jobs.csv and test/input/3jobs_zeros.csv don't have the same runtime (+/- 1sec). It is unusual, as runtime are normally an input of the simulation.
+      warnings.warn(f"Some jobs in {file1} and {file2} don't have the same runtime (+/- 1sec). It is unusual, as runtime are normally an input of the simulation.")
+    /home/mael/ownCloud/workspace/batmen-tools/distance_batsim_output.py:114: RuntimeWarning: divide by zero encountered in double_scalars
+      dist[f]["normalized_euclidean"] = eucl / np.sqrt(sum_runtimes1 * sum_runtimes2)
    {
        "submission_time": {
            "euclidean": 50.0,
-            "normalized_euclidean": null,
+            "normalized_euclidean": Infinity,
            "lateness": -70.0
        },
        "starting_time": {
            "euclidean": 50.0,
-            "normalized_euclidean": null,
+            "normalized_euclidean": Infinity,
            "lateness": -70.0
        },
        "finish_time": {
            "euclidean": 100.0,
-            "normalized_euclidean": null,
+            "normalized_euclidean": Infinity,
            "lateness": -140.0
        }
    }

 %% Cell type:code id: tags:

 ``` python
 !python3 distance_batsim_output.py {mc_10days_a60} {mc_10days_m60} --all
 ```

 %% Output

    {
        "submission_time": {
            "euclidean": 172.1394783308001,
-            "normalized_euclidean": 2.739806304756353e-10,
+            "normalized_euclidean": 1.2942984329407551e-05,
            "lateness": 2254.0
        },
        "starting_time": {
            "euclidean": 172.1394783308001,
-            "normalized_euclidean": 2.739806304756353e-10,
+            "normalized_euclidean": 1.2942984329407551e-05,
            "lateness": 2254.0
        },
        "finish_time": {
            "euclidean": 172.1394783308001,
-            "normalized_euclidean": 2.3852456580198333e-10,
+            "normalized_euclidean": 1.2942984329407551e-05,
            "lateness": 2254.0
        }
    }

 %% Cell type:code id: tags:

 ``` python
 !python3 distance_batsim_output.py {mc_10days_a60} {mc_10days_rigid} --all
 ```

 %% Output

    {
        "submission_time": {
            "euclidean": 241.15762480170517,
-            "normalized_euclidean": 5.377222025784075e-10,
+            "normalized_euclidean": 1.813216587526566e-05,
            "lateness": 3311.0
        },
        "starting_time": {
            "euclidean": 241.15762480170517,
-            "normalized_euclidean": 5.377222025784075e-10,
+            "normalized_euclidean": 1.813216587526566e-05,
            "lateness": 3311.0
        },
        "finish_time": {
            "euclidean": 255.14897609044016,
-            "normalized_euclidean": 5.240304046302593e-10,
+            "normalized_euclidean": 1.91841479662116e-05,
            "lateness": 3633.0
        }
    }

 %% Cell type:code id: tags:

 ``` python
 ```