Skip to content

Commit

Permalink
Merge pull request #41 from acisops/Use_CR_Backstop_Files
Browse files Browse the repository at this point in the history 
Update for SCS-106 change and new power commands
  • Loading branch information
@Gregg140
Gregg140 authored Oct 18, 2021
2 parents 5d1c118 + 726f74a commit 65639e9
Show file tree
Hide file tree
Showing 3 changed files with 80 additions and 154 deletions.
74 changes: 43 additions & 31 deletions acis_thermal_check/main.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -30,13 +30,11 @@
from astropy.table import Table
import ska_helpers


op_map = {"greater": ">",
"greater_equal": ">=",
"less": "<",
"less_equal": "<="}


class ACISThermalCheck(object):
r"""
ACISThermalCheck class for making thermal model predictions
Expand Down Expand Up @@ -109,6 +107,7 @@ def __init__(self, msid, name, validation_limits, hist_limit,
hist_ops = ["greater_equal"]*len(hist_limit)
self.hist_ops = hist_ops
self.perigee_passages = []
self.write_pickle = False

def _handle_limits(self):
from yaml import load, Loader
Expand Down Expand Up @@ -147,6 +146,10 @@ def run(self, args, override_limits=None):

proc = self._setup_proc_and_logger(args)

# If args.run_start is not none, write validation and prediction
# data to a pickle later
self.write_pickle = args.run_start is not None

# This allows one to override the planning and yellow limits
# for a particular model run. THIS SHOULD ONLY BE USED FOR
# TESTING PURPOSES.
Expand All @@ -161,8 +164,8 @@ def run(self, args, override_limits=None):
# stored in state_builder or punt by using NOW and None for
# tstart and tstop.
is_weekly_load = args.backstop_file is not None
tstart, tstop, tnow = self._determine_times(args.run_start,
is_weekly_load)
tstart, tstop, t_run_start = self._determine_times(args.run_start,
is_weekly_load)

# Store off the start date, and, if you have it, the
# stop date in proc
Expand All @@ -171,34 +174,32 @@ def run(self, args, override_limits=None):
proc["datestop"] = CxoTime(tstop).date

# Get the telemetry values which will be used
# for prediction and validation. Args default value is 21 days.
tlm = self.get_telem_values(min(tstart, tnow), days=args.days)
# for prediction and validation. Validation runs
# begin "args.days" before the start of the prediction
# run. "args.days" default value is 21 days.
tlm = self.get_telem_values(min(tstart, t_run_start), days=args.days)

# make predictions on a backstop file if defined
if args.backstop_file is not None:
pred = self.make_week_predict(tstart, tstop, tlm, args.T_init,
args.model_spec, args.outdir)
args.model_spec, args.outdir, args.run_start)
else:
pred = defaultdict(lambda: None)

# Validation

if not args.pred_only:

# Make the validation plots
plots_validation = self.make_validation_plots(tlm, args.model_spec,
args.outdir,
args.run_start)
args.outdir)

proc["op"] = [op_map[op] for op in self.hist_ops]

# Determine violations of temperature validation
valid_viols = self.make_validation_viols(plots_validation)
if len(valid_viols) > 0:
mylog.info('validation warning(s) in output at %s' % args.outdir)

else:

valid_viols = defaultdict(lambda: None)
plots_validation = defaultdict(lambda: None)

Expand Down Expand Up @@ -234,7 +235,7 @@ def get_ephemeris(self, start, stop, times):
times)
return ephem

def get_states(self, tlm, T_init):
def get_states(self, tlm, T_init, run_start_val):
"""
Call the state builder to get the commanded states and
determine the initial temperature.
Expand All @@ -246,10 +247,14 @@ def get_states(self, tlm, T_init):
T_init : float
The initial temperature of the model prediction. If None, an
initial value will be constructed from telemetry.
Calculated values: tbegin DOY string used throughout the model to
indicate when to stop backchaining as well
"""
# The -5 here has us back off from the last telemetry
# reading just a bit

# --run_start not specified. Back off -5 from the last telemetry
tbegin = CxoTime(tlm['date'][-5]).date

# Call the overloaded state_builder method to assemble states
# and define a state0
states, state0 = self.state_builder.get_prediction_states(tbegin)
Expand All @@ -269,7 +274,7 @@ def get_states(self, tlm, T_init):
return states, state0

def make_week_predict(self, tstart, tstop, tlm, T_init, model_spec,
outdir):
outdir, run_start_val):
"""
Parameters
----------
Expand All @@ -292,7 +297,7 @@ def make_week_predict(self, tstart, tstop, tlm, T_init, model_spec,
mylog.info('Calculating %s thermal model' % self.name.upper())

# Get commanded states and set initial temperature
states, state0 = self.get_states(tlm, T_init)
states, state0 = self.get_states(tlm, T_init, run_start_val)

# calc_model actually does the model calculation by running
# model-specific code.
Expand All @@ -316,8 +321,10 @@ def make_week_predict(self, tstart, tstop, tlm, T_init, model_spec,

# make_prediction_viols determines the violations and prints them out
viols = self.make_prediction_viols(temps, tstart)

# write_states writes the commanded states to states.dat
self.write_states(outdir, states)

# write_temps writes the temperatures to temperatures.dat
self.write_temps(outdir, model.times, temps)

Expand Down Expand Up @@ -578,14 +585,14 @@ def write_temps(self, outdir, times, temps):
temp_table[self.msid].format = '%.2f'
temp_table.write(outfile, format='ascii', delimiter='\t', overwrite=True)

def _gather_perigee(self, run_start, load_start):
def _gather_perigee(self, plot_start, load_start):
import glob
# The first step is to build a list of all the perigee passages.

# Gather the perigee passages that occur from the
# beginning of the model run up to the start of the load
# from kadi
rzs = events.rad_zones.filter(run_start, load_start)
rzs = events.rad_zones.filter(plot_start, load_start)
for rz in rzs:
self.perigee_passages.append([rz.start, rz.perigee])
for rz in rzs:
Expand Down Expand Up @@ -798,7 +805,7 @@ def get_histogram_mask(self, tlm, limits):
masks.append(mask)
return masks

def make_validation_plots(self, tlm, model_spec, outdir, run_start):
def make_validation_plots(self, tlm, model_spec, outdir):
"""
Make validation output plots by running the thermal model from a
time in the past forward to the present and compare it to real
Expand All @@ -812,8 +819,6 @@ def make_validation_plots(self, tlm, model_spec, outdir, run_start):
The path to the thermal model specification.
outdir : string
The directory to write outputs to.
run_start : string
The starting date/time of the run.
"""
import pickle

Expand Down Expand Up @@ -1067,10 +1072,13 @@ def make_validation_plots(self, tlm, model_spec, outdir, run_start):
f.write(quant_table)
f.close()

# If run_start is specified this is likely for regression testing
# or other debugging. In this case write out the full predicted and
# telemetered dataset as a pickle.
if run_start:
# self.write_pickle is set to the value of True or False based upon the
# value of the command line argument: --run-start. --run-start can be
# either a DOY date string or None (if the argument
# was not specified). If a DOY date, this model run is likely for regression
# testing or other debugging. In that case write out the full
# predicted and telemetered dataset as a pickle.
if self.write_pickle:
filename = os.path.join(outdir, 'validation_data.pkl')
mylog.info('Writing validation data %s' % filename)
f = open(filename, 'wb')
Expand Down Expand Up @@ -1207,7 +1215,8 @@ def _setup_proc_and_logger(self, args):

def _determine_times(self, run_start, is_weekly_load):
"""
Determine the start and stop times
Determine the start and stop times, as well as the
"CXC time" corresponding to the run_start argument.
Parameters
----------
Expand All @@ -1216,7 +1225,9 @@ def _determine_times(self, run_start, is_weekly_load):
is_weekly_load : boolean
Whether or not this is a weekly load.
"""
tnow = CxoTime(run_start).secs
# Note: if run_start is None, then it defaults to
# whatever the time is RIGHT NOW
t_run_start = CxoTime(run_start).secs
# Get tstart, tstop, commands from state builder
if is_weekly_load:
# If we are running a model for a particular load,
Expand All @@ -1225,12 +1236,13 @@ def _determine_times(self, run_start, is_weekly_load):
tstart = self.state_builder.tstart
tstop = self.state_builder.tstop
else:
# Otherwise, the start time for the run is whatever is in
# Otherwise, we are just doing a validation run, and
# the start time for the run is whatever is in
# args.run_start
tstart = tnow
tstart = t_run_start
tstop = None

return tstart, tstop, tnow
return tstart, tstop, t_run_start

def get_telem_values(self, tstart, days=14):
"""
Expand Down
Loading

0 comments on commit 65639e9

Please sign in to comment.