[PATCH] osmo-gsm-tester[master]: ms: Create a cumulative distribution function class

This is merely a historical archive of years 2008-2021, before the migration to mailman3.

A maintained and still updated list archive can be found at https://lists.osmocom.org/hyperkitty/list/gerrit-log@lists.osmocom.org/.

Holger Freyther gerrit-no-reply at lists.osmocom.org
Mon Feb 26 21:35:06 UTC 2018


Hello Pau Espin Pedrol, Harald Welte, Jenkins Builder,

I'd like you to reexamine a change.  Please visit

    https://gerrit.osmocom.org/6230

to look at the new patch set (#4).

ms: Create a cumulative distribution function class

We are using the CDF to decide which percentage of the jobs should
be running at a given point. The x-axis is time and the y-axis the
percentage of how many jobs should be running.

There are three functions to do this. The first one is a constant
which would result in everything being started right now, one to
start them linearly and the last (formula from Qt/3rdparty) to first
accelerate and decelerate slowly.

Change-Id: I9e3064f4c3c4c7af5d3491f850090516e541f4d3
---
A src/osmo_ms_driver/__init__.py
A src/osmo_ms_driver/cdf.py
2 files changed, 252 insertions(+), 0 deletions(-)


  git pull ssh://gerrit.osmocom.org:29418/osmo-gsm-tester refs/changes/30/6230/4

diff --git a/src/osmo_ms_driver/__init__.py b/src/osmo_ms_driver/__init__.py
new file mode 100644
index 0000000..d3c1590
--- /dev/null
+++ b/src/osmo_ms_driver/__init__.py
@@ -0,0 +1,29 @@
+# osmo_gsm_tester: automated cellular network hardware tests
+#
+# Copyright (C) 2016-2017 by sysmocom - s.f.m.c. GmbH
+#
+# Authors: D. Lazlo Sitzer <dlsitzer at sysmocom.de>
+#          Neels Hofmeyr <neels at hofmeyr.de>
+#
+# This program is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as
+# published by the Free Software Foundation, either version 3 of the
+# License, or (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+__version__ = 'UNKNOWN'
+
+try:
+    from ._version import _version
+    __version__ = _version
+except:
+    pass
+
+# vim: expandtab tabstop=4 shiftwidth=4
diff --git a/src/osmo_ms_driver/cdf.py b/src/osmo_ms_driver/cdf.py
new file mode 100644
index 0000000..d5809e1
--- /dev/null
+++ b/src/osmo_ms_driver/cdf.py
@@ -0,0 +1,223 @@
+# osmo_ms_driver: A cumululative distribution function class.
+# Help to start processes over time.
+#
+# Copyright (C) 2018 by Holger Hans Peter Freyther
+#
+# This program is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as
+# published by the Free Software Foundation, either version 3 of the
+# License, or (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+
+from datetime import timedelta
+
+class DistributionFunctionHandler(object):
+    """
+    The goal is to start n "mobile" processes. We like to see some
+    conflicts (RACH bursts being ignored) but starting n processes
+    at the same time is not a realistic model.
+    We use the concept of cumulative distribution function here. On
+    the x-axis we have time (maybe in steps of 10ms) and on the
+    y-axis we have the percentage (from 0.0 to 1.0) of how many
+    processes should run at the given time.
+    """
+
+    def __init__(self, step, duration, fun):
+        self._step = step
+        self._fun = fun
+        self._x = 0.0
+        self._y = self._fun(self._x)
+        self._target = 1.0
+        self._duration = duration
+
+    def step_size(self):
+        return self._step
+
+    def set_target(self, scale):
+        """
+        Scale the percentage to the target value..
+        """
+        self._target = scale
+
+    def is_done(self):
+        return self._y >= 1.0
+
+    def current_value(self):
+        return self._y
+
+    def current_scaled_value(self):
+        return self._y * self._target
+
+    def step_once(self):
+        self._x = self._x + self._step.total_seconds()
+        self._y = self._fun(self._x)
+
+    def duration(self):
+        return self._duration
+
+
+def immediate(step_size=timedelta(milliseconds=20)):
+    """
+    Reaches 100% at the first step.
+
+    Example:
+    >>> a = immediate()
+    >>> a.is_done()
+    True
+    >>> a.current_value()
+    1
+    """
+    duration = timedelta(seconds=0)
+    return DistributionFunctionHandler(step_size, duration, lambda x: 1)
+
+def linear_with_slope(slope, duration, step_size=timedelta(milliseconds=20)):
+    """
+    Use the slope and step size you want
+    """
+    return DistributionFunctionHandler(step_size, duration, lambda x: slope*x)
+
+def linear_with_duration(duration, step_size=timedelta(milliseconds=20)):
+    """
+    Linear progression that reaches 100% after duration.total_seconds()
+    >>> a = linear_with_duration(timedelta(seconds=10), step_size=timedelta(seconds=2))
+    >>> a.is_done()
+    False
+    >>> a.current_value()
+    0.0
+    >>> a.step_once()
+    >>> a.current_value()
+    0.2
+    >>> a.step_once()
+    >>> a.current_value()
+    0.4
+    >>> a.step_once()
+    >>> a.current_value()
+    0.6...
+    >>> a.step_once()
+    >>> a.current_value()
+    0.8...
+    >>> a.is_done()
+    False
+    >>> a.step_once()
+    >>> a.current_value()
+    1.0...
+    >>> a.is_done()
+    True
+
+    >>> a = linear_with_duration(timedelta(seconds=10), step_size=timedelta(seconds=2))
+    >>> a.set_target(1000)
+    >>> a.is_done()
+    False
+    >>> a.current_value()
+    0.0
+    >>> a.current_scaled_value()
+    0.0
+    >>> a.step_once()
+    >>> a.current_value()
+    0.2
+    >>> a.current_scaled_value()
+    200.0
+    >>> a.step_once()
+    >>> a.current_value()
+    0.4
+    >>> a.current_scaled_value()
+    400.0
+    >>> a.step_once()
+    >>> a.current_value()
+    0.6...
+    >>> a.current_scaled_value()
+    600.0...
+    >>> a.step_once()
+    >>> a.current_value()
+    0.8...
+    >>> a.current_scaled_value()
+    800.0...
+    >>> a.is_done()
+    False
+    >>> a.step_once()
+    >>> a.current_value()
+    1.0...
+    >>> a.current_scaled_value()
+    1000.0...
+    >>> a.is_done()
+    True
+    """
+    slope = 1.0/duration.total_seconds()
+    return linear_with_slope(slope, duration, step_size)
+
+def _in_out(x):
+    """
+    Internal in/out function inspired by Qt
+
+    >>> _in_out(0.5)
+    0.5...
+    >>> _in_out(0.75)
+    0.875...
+    >>> _in_out(0.8)
+    0.92...
+    >>> _in_out(0.85)
+    0.955...
+    >>> _in_out(1.0)
+    1.0...
+    """
+    assert x <= 1.0
+    # Needs to be between 0..1 and increase first
+    if x < 0.5:
+        return (x*x) * 2
+    # deaccelerate now. in_out(0.5) == 0.5, in_out(1.0) == 1.0
+    x = x * 2 - 1
+    return -0.5 * (x*(x-2)- 1)
+
+def ease_in_out_duration(duration, step_size=timedelta(milliseconds=20)):
+    """
+    Example invocation
+
+    >>> a = ease_in_out_duration(duration=timedelta(seconds=20), step_size=timedelta(seconds=5))
+    >>> a.is_done()
+    False
+    >>> a.current_value()
+    0.0
+
+    >>> a.step_once()
+    >>> a._x
+    5.0
+    >>> a.current_value()
+    0.125
+    >>> a.is_done()
+    False
+
+    >>> a.step_once()
+    >>> a._x
+    10.0
+    >>> a.current_value()
+    0.5
+    >>> a.is_done()
+    False
+
+    >>> a.step_once()
+    >>> a._x
+    15.0
+    >>> a.current_value()
+    0.875
+    >>> a.is_done()
+    False
+
+    >>> a.step_once()
+    >>> a._x
+    20.0
+    >>> a.current_value()
+    1.0
+    >>> a.is_done()
+    True
+    """
+    scale = 1.0/duration.total_seconds()
+    return DistributionFunctionHandler(step_size, duration,
+                                        lambda x: _in_out(x*scale))

-- 
To view, visit https://gerrit.osmocom.org/6230
To unsubscribe, visit https://gerrit.osmocom.org/settings

Gerrit-MessageType: newpatchset
Gerrit-Change-Id: I9e3064f4c3c4c7af5d3491f850090516e541f4d3
Gerrit-PatchSet: 4
Gerrit-Project: osmo-gsm-tester
Gerrit-Branch: master
Gerrit-Owner: Holger Freyther <holger at freyther.de>
Gerrit-Reviewer: Harald Welte <laforge at gnumonks.org>
Gerrit-Reviewer: Holger Freyther <holger at freyther.de>
Gerrit-Reviewer: Jenkins Builder
Gerrit-Reviewer: Pau Espin Pedrol <pespin at sysmocom.de>
Gerrit-Reviewer: Vadim Yanitskiy <axilirator at gmail.com>



More information about the gerrit-log mailing list