# Copyright 2015-2016 ARM Limited
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
#
import pandas as pd
import trappy
from utils_tests import TestBART
from bart.common.signal import SignalCompare
import numpy as np
class TestSignalCompare(TestBART):
def __init__(self, *args, **kwargs):
super(TestSignalCompare, self).__init__(*args, **kwargs)
def test_conditional_compare(self):
"""Test conditional_compare"""
# Refer to the example in
# bart.common.signal.SignalCompare.conditional_compare
# doc-strings which explains the calculation for the
# data set below
A = [0, 0, 0, 3, 3, 0, 0, 0]
B = [0, 0, 2, 2, 2, 2, 1, 1]
trace = trappy.BareTrace()
df = pd.DataFrame({"A": A, "B": B})
trace.add_parsed_event("event", df)
s = SignalCompare(trace, "event:A", "event:B")
expected = (1.5, 2.0 / 7)
self.assertEqual(
s.conditional_compare(
"event:A > event:B",
method="rect"),
expected)
def test_get_overshoot(self):
"""Test get_overshoot"""
A = [0, 0, 0, 3, 3, 0, 0, 0]
B = [0, 0, 2, 2, 2, 2, 1, 1]
trace = trappy.BareTrace()
df = pd.DataFrame({"A": A, "B": B})
trace.add_parsed_event("event", df)
s = SignalCompare(trace, "event:A", "event:B")
expected = (1.5, 2.0 / 7)
self.assertEqual(
s.get_overshoot(method="rect"),
expected)
A = [0, 0, 0, 1, 1, 0, 0, 0]
B = [0, 0, 2, 2, 2, 2, 1, 1]
df = pd.DataFrame({"A": A, "B": B})
trace.event.data_frame = df
s = SignalCompare(trace, "event:A", "event:B")
expected = (float("nan"), 0.0)
result = s.get_overshoot(method="rect")
self.assertTrue(np.isnan(result[0]))
self.assertEqual(result[1], expected[1])
def test_get_undershoot(self):
"""Test get_undershoot"""
A = [0, 0, 0, 1, 1, 1, 1, 1]
B = [2, 2, 2, 2, 2, 2, 2, 2]
trace = trappy.BareTrace()
df = pd.DataFrame({"A": A, "B": B})
trace.add_parsed_event("event", df)
s = SignalCompare(trace, "event:A", "event:B")
expected = (4.0 / 14.0, 1.0)
self.assertEqual(
s.get_undershoot(method="rect"),
expected)
A = [3, 3, 3, 3, 3, 3, 3, 3]
B = [2, 2, 2, 2, 2, 2, 1, 1]
df = pd.DataFrame({"A": A, "B": B})
trace.event.data_frame = df
s = SignalCompare(trace, "event:A", "event:B")
expected = (float("nan"), 0.0)
result = s.get_undershoot(method="rect")
self.assertTrue(np.isnan(result[0]))
self.assertEqual(result[1], expected[1])