From adf1490fc52c82c631e29922670d3c0765b73551 Mon Sep 17 00:00:00 2001
From: Christian Huitema <huitema@huitema.net>
Date: Mon, 15 Apr 2024 12:17:26 -0700
Subject: [PATCH] Save sampler script

---
 imrs/imrs_sampler.py | 56 +++++++++++++++++++++++++++++
 src/imrs_total.py    | 86 ++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 142 insertions(+)
 create mode 100644 imrs/imrs_sampler.py
 create mode 100644 src/imrs_total.py

diff --git a/imrs/imrs_sampler.py b/imrs/imrs_sampler.py
new file mode 100644
index 0000000..a600914
--- /dev/null
+++ b/imrs/imrs_sampler.py
@@ -0,0 +1,56 @@
+#
+# This script will try to build a sample of the input file.
+# The purpose of the sample is, get a realistic test file
+# that is small enough for iterative development, measures,
+# etc., yet big enough to obtain statistically significant
+# results.
+#
+# Usage: imrs_sample.py <input_file> <sampling rate in %> <output_file>
+#
+
+import sys
+import traceback
+import random
+import time
+import concurrent.futures
+import os
+from os import listdir
+from os.path import isfile, isdir, join
+
+
+# main
+
+if len(sys.argv) != 4:
+    print("Usage: imrs_sample.py <input_file> <sampling rate in %> <output_file>")
+    exit(1)
+file_name_in = sys.argv[1]
+rate_text = sys.argv[2]
+file_name_out = sys.argv[3]
+sampling_rate = 0
+if not rate_text.endswith("%"):
+    print("sampling rate should be e.g. 5\%, 0.1%, not " + rate_text)
+    exit(1)
+try:
+    rate_percent = float(rate_text[:-1])
+    sampling_rate = rate_percent/100.0
+except Exception as e:
+    traceback.print_exc()
+    print("Cannot parse <" + rate_percent  + ">\nException: " + str(e))
+    exit(1)
+
+nb_lines_in = 0
+nb_lines_out = 0
+with open(file_name_out,"wt") as F_OUT:
+    for line in open(file_name_in, "rt"):
+        nb_lines_in += 1
+        if random.random() < sampling_rate:
+            F_OUT.write(line)
+            nb_lines_out += 1
+
+if nb_lines_in == 0:
+    print("Input file " + file_name_in + " is empty.")
+else:
+    print("Input file " + file_name_in + ": " + str(nb_lines_in) + " lines")
+    print("Output file " + file_name_out + ": " + str(nb_lines_out) + " lines")
+    print("Sampling rate requested: " + str(sampling_rate))
+    print("Sampling rate actual: " + str(nb_lines_out/nb_lines_in))
diff --git a/src/imrs_total.py b/src/imrs_total.py
new file mode 100644
index 0000000..8ac88cf
--- /dev/null
+++ b/src/imrs_total.py
@@ -0,0 +1,86 @@
+#!/usr/bin/python
+# coding=utf-8
+#
+# This script organizes the sum of IMRS resolver data per
+# cluster for the month. The results are collected in
+# the folder  ~/ipstats/cluster, with one subfolder
+# per cluster, using names like ~/ipstats/cluster/us-lax/.
+# For each cluster, the script compute a single file
+# such as ~/ipstats/cluster/us-lax.202403.csv, containing
+# the aggregated statistics for the whole month.
+#
+
+import sys
+import traceback
+import random
+import time
+import concurrent.futures
+import os
+from os import listdir
+from os.path import isfile, isdir, join
+
+def check_or_create_dir(dir_path):
+    if not isdir(dir_path):
+        try:
+            os.mkdir(dir_path)
+        except Exception as e:
+            traceback.print_exc()
+            print("Cannot create <" + dir_path  + ">\nException: " + str(e))
+            return False
+    return True
+
+# main
+if len(sys.argv) < 4 or len(sys.argv) > 5 or \
+    (len(sys.argv) == 5 and sys.argv[4] != "debug"):
+    print("Usage: imrs_monthly <ipstats_folder> <yyyymm> <ithitool> [\"debug\"]")
+    print("There are just " + str(len(sys.argv)) + " arguments.")
+    exit (1)
+ipstats_folder = sys.argv[1]
+month = sys.argv[2]
+ithitool = sys.argv[3]
+do_debug = len(sys.argv) == 5
+
+print("Writing monthly per custom clusters aggregates for: " + ipstats_folder)
+try:
+    # Look at every cluster under the "clusters" folder
+    monthly_folder = join(ipstats_folder, "monthly")
+    tmp_folder = join(ipstats_folder, "tmp")
+    monthly_list = listdir(monthly_folder)
+    if check_or_create_dir(monthly_folder) and \
+       check_or_create_dir(tmp_folder):
+        tmp_file_name = join(tmp_folder, month + ".txt")
+        with open(tmp_file_name, "wt") as F:
+            # check that this is a cluster, and not some other file
+            # Watch for: cluster_id + "." + month + "-" + "ipstats.csv"
+            monthly_file_end = month + "-" + "ipstats.csv"
+            for monthly_file in monthly_list:
+                monthly_path = join(monthly_folder, monthly_file)
+                if len(monthly_file) > 7 and \
+                    monthly_file[2] == "-" and \
+                    monthly_file[6] == "." and \
+                    monthly_file.endswith(monthly_file_end):
+                    F.write(monthly_path +"\n")
+                    if do_debug:
+                        print("Adding: " + monthly_file)
+                elif do_debug:
+                    print("Not a monthly file: " + monthly_path)
+        total_file = "total-" + month +  "-" + "ipstats.csv"
+        total_path = join(ipstats_folder, total_file)
+        merge_cmd = ithitool + ' -I ' + total_path + " " + tmp_file_name
+        if do_debug:
+            print("Running: " + merge_cmd)
+            sys.stdout.flush()
+        cmd_ret = os.system(merge_cmd)
+        if cmd_ret == 0:
+            if do_debug:
+                print(total_file + ": computed.")
+        else:
+            print(report_name + ": computation failed, error:" + str(cmd_ret))
+except Exception as exc:
+   traceback.print_exc()
+   print('\nCode generated an exception: %s' % (exc))
+
+
+
+
+