summaryrefslogtreecommitdiff
path: root/lib/dfatool.py
diff options
context:
space:
mode:
authorDaniel Friesel <daniel.friesel@uos.de>2019-09-23 15:19:11 +0200
committerDaniel Friesel <daniel.friesel@uos.de>2019-09-23 15:19:11 +0200
commit3f631ac30c42751daaca6255da02d5b76de077c9 (patch)
treec2572985909bac993e8e80a36161eced0a7df38d /lib/dfatool.py
parentea89af21e3d9cc9b8ff856105aa36698881ef565 (diff)
Fix model generation on legacy data format
A bug introduced by 3b624b1bb4e3614d5befbe5b367d1506dad3c933 caused traces to be duplicated in a not entirely predictable manner, occasionally affecting model data.
Diffstat (limited to 'lib/dfatool.py')
-rwxr-xr-xlib/dfatool.py53
1 files changed, 39 insertions, 14 deletions
diff --git a/lib/dfatool.py b/lib/dfatool.py
index c81ae0a..68146ae 100755
--- a/lib/dfatool.py
+++ b/lib/dfatool.py
@@ -360,12 +360,14 @@ def _preprocess_measurement(measurement):
'triggers' : len(trigidx),
'first_trig' : trigidx[0] * 10,
'calibration' : caldata,
- 'trace' : mim.analyze_states(charges, trigidx, vcalfunc),
- 'expected_trace' : measurement['expected_trace'],
+ 'energy_trace' : mim.analyze_states(charges, trigidx, vcalfunc),
'has_mimosa_error' : mim.is_error,
'mimosa_errors' : mim.errors,
}
+ if 'expected_trace' in measurement:
+ processed_data['expected_trace'] = measurement['expected_trace']
+
return processed_data
class ParamStats:
@@ -656,7 +658,9 @@ class RawData:
'triggers' : len(trigidx),
'first_trig' : trigidx[0] * 10,
'calibration' : caldata,
- 'trace' : mim.analyze_states(charges, trigidx, vcalfunc)
+ 'energy_trace' : mim.analyze_states(charges, trigidx, vcalfunc)
+ A sequence of unnamed, unparameterized states and transitions with
+ energy and timing data
'expected_trace' : trace from PTA DFS (with parameter data)
mim.analyze_states returns a list of (alternating) states and transitions.
Each element is a dict containing:
@@ -674,7 +678,10 @@ class RawData:
- uW_mean_delta_next: Differenz zwischen uW_mean und uW_mean des Folgezustands
"""
setup = self.setup_by_fileno[processed_data['fileno']]
- traces = processed_data['expected_trace']
+ if 'expected_trace' in processed_data:
+ traces = processed_data['expected_trace']
+ else:
+ traces = self.traces_by_fileno[processed_data['fileno']]
state_duration = setup['state_duration']
# Check MIMOSA error
@@ -703,7 +710,7 @@ class RawData:
online_datapoints.append((run_idx, trace_part_idx))
for offline_idx, online_ref in enumerate(online_datapoints):
online_run_idx, online_trace_part_idx = online_ref
- offline_trace_part = processed_data['trace'][offline_idx]
+ offline_trace_part = processed_data['energy_trace'][offline_idx]
online_trace_part = traces[online_run_idx]['trace'][online_trace_part_idx]
if self._parameter_names == None:
@@ -763,14 +770,21 @@ class RawData:
return True
def _merge_online_and_offline(self, measurement):
+ # Edits self.traces_by_fileno[measurement['fileno']][*]['trace'][*]['offline']
+ # and self.traces_by_fileno[measurement['fileno']][*]['trace'][*]['offline_aggregates'] in place
+ # (appends data from measurement['energy_trace'])
+ # If measurement['expected_trace'] exists, it is edited in place instead
online_datapoints = []
- traces = measurement['expected_trace'].copy()
+ if 'expected_trace' in measurement:
+ traces = measurement['expected_trace']
+ else:
+ traces = self.traces_by_fileno[measurement['fileno']]
for run_idx, run in enumerate(traces):
for trace_part_idx in range(len(run['trace'])):
online_datapoints.append((run_idx, trace_part_idx))
for offline_idx, online_ref in enumerate(online_datapoints):
online_run_idx, online_trace_part_idx = online_ref
- offline_trace_part = measurement['trace'][offline_idx]
+ offline_trace_part = measurement['energy_trace'][offline_idx]
online_trace_part = traces[online_run_idx]['trace'][online_trace_part_idx]
if not 'offline' in online_trace_part:
@@ -823,7 +837,6 @@ class RawData:
offline_trace_part['uW_mean_delta_next'] * (offline_trace_part['us'] - 20))
online_trace_part['offline_aggregates']['timeout'].append(
offline_trace_part['timeout'])
- return traces
def _concatenate_traces(self, list_of_traces):
trace_output = list()
@@ -914,7 +927,6 @@ class RawData:
'fileno' : i,
'info' : member,
'setup' : self.setup_by_fileno[i],
- 'expected_trace' : self.traces_by_fileno[i],
})
elif version == 1:
@@ -922,6 +934,8 @@ class RawData:
traces_by_file = list()
mim_files_by_file = list()
with tarfile.open(filename) as tf:
+ # Relies on generate-dfa-benchmark placing the .mim files
+ # in the order they were created (i.e., lexically sorted)
for member in tf.getmembers():
_, extension = os.path.splitext(member.name)
if extension == '.mim':
@@ -931,8 +945,16 @@ class RawData:
})
elif extension == '.json':
ptalog = json.load(tf.extractfile(member))
+
+ # ptalog['traces'] is a list of lists.
+ # The first level corresponds to the individual .mim files:
+ # ptalog['traces'][0] contains all traces belonging to the
+ # first .mim file in the archive.
+ # The second level holds the individual runs in this
+ # sub-benchmark, so ptalog['traces'][0][0] is the first
+ # run, ptalog['traces'][0][1] the second, and so on
+
traces_by_file.extend(ptalog['traces'])
- self.traces_by_fileno.append(self._concatenate_traces(traces_by_file))
self.setup_by_fileno.append({
'mimosa_voltage' : ptalog['configs'][0]['voltage'],
'mimosa_shunt' : ptalog['configs'][0]['shunt'],
@@ -953,16 +975,16 @@ class RawData:
if version == 0:
# Strip the last state (it is not part of the scheduled measurement)
- measurement['trace'].pop()
+ measurement['energy_trace'].pop()
repeat = 0
elif version == 1:
# The first online measurement is the UNINITIALIZED state. In v1,
# it is not part of the expected PTA trace -> remove it.
- measurement['trace'].pop(0)
+ measurement['energy_trace'].pop(0)
repeat = ptalog['opt']['repeat']
if self._measurement_is_valid_01(measurement, repeat):
- valid_traces.append(self._merge_online_and_offline(measurement))
+ self._merge_online_and_offline(measurement)
num_valid += 1
else:
vprint(self.verbose, '[W] Skipping {ar:s}/{m:s}: {e:s}'.format(
@@ -972,7 +994,10 @@ class RawData:
vprint(self.verbose, '[I] {num_valid:d}/{num_total:d} measurements are valid'.format(
num_valid = num_valid,
num_total = len(measurements)))
- self.traces = self._concatenate_traces(valid_traces)
+ if version == 0:
+ self.traces = self._concatenate_traces(self.traces_by_fileno)
+ elif version == 1:
+ self.traces = self._concatenate_traces(map(lambda x: x['expected_trace'], measurements))
self.preprocessing_stats = {
'num_runs' : len(measurements),
'num_valid' : num_valid