def test_genvpdbs():
lc_temp_dir = TEMP_DIR + LIGHT_CURVES_DIR
db_temp_dir = TEMP_DIR + DB_DIR
# Create 10 vantage point database files
genvpdbs.create_vpdbs(10, lc_temp_dir, db_temp_dir)
# With vantage points created, we should have fully created dataset
assert (simsearch.need_to_rebuild(lc_temp_dir, db_temp_dir) == False)
# Load vantage points back in from disk
vp_dict = simsearch.load_vp_lcs(db_temp_dir, lc_temp_dir)
assert (len(vp_dict) == 10)
clear_dir(TEMP_DIR, recreate=False)
def test_cmd_line_util():
os.chdir('cs207project/tsrbtreedb')
_ = simsearchutil.cmd_line_util()
old_sys_argv = sys.argv
sys.argv = old_sys_argv + ["-r", "-d"]
_ = simsearchutil.cmd_line_util()
old_sys_argv = sys.argv
sys.argv = old_sys_argv + ["-h"]
_ = simsearchutil.cmd_line_util()
clear_dir(LIGHT_CURVES_DIR, recreate=False)
clear_dir(DB_DIR, recreate=False)
os.chdir(os.pardir)
os.chdir(os.pardir)
def test_makelcs():
lc_temp_dir = TEMP_DIR + LIGHT_CURVES_DIR
db_temp_dir = TEMP_DIR + DB_DIR
# Before we've generated anything, confirm that our rebuild detector method properly detects that we need to rebuild
assert (simsearch.need_to_rebuild(lc_temp_dir, db_temp_dir) == True)
makelcs.make_lcs_wfm(5, lc_temp_dir)
assert (simsearch.need_to_rebuild(lc_temp_dir, db_temp_dir) == True)
clear_dir(TEMP_DIR, recreate=False)
# Generate 100 light curves in temp dir
makelcs.make_lcs_wfm(100, lc_temp_dir)
assert (simsearch.need_to_rebuild(lc_temp_dir, db_temp_dir) == True)
# load that back in from disk
lc_dict = genvpdbs.load_ts_fsm(lc_temp_dir)
assert (len(lc_dict) == 100)
def create_vpdbs(n, lc_dir, db_dir):
"""
Create Vantage point databases in parallel
Executes functions above:
(1) Creates timeseries_dict from time series files on disk
(2) Picks 20 vantage points at random
(3) Using a process poll for parallelization, calculates kernel distance between vantage points and generated time series
(4) Saves kernel distance indexes to disk as red-black tree databases
"""
print("Creating %d vantage point dbs" % n, end="")
timeseries_dict = load_ts_fsm(lc_dir)
vantage_points = pick_vantage_points(timeseries_dict, n)
clear_dir(db_dir)
# This is in elegant, but needed to use the map function below where
# we can effectually only pass one augment to the worker process
vp_tuples = [(vp, timeseries_dict, db_dir) for vp in vantage_points]
# build vantage point dbs in parallel (up to the number of processes on your machine)
with ProcessPoolExecutor() as pool:
results = pool.map(save_vp_dbs, vp_tuples)
print("....................Done.")
def test_setup_temp_dir():
"""Clear existing temp dir and change working dir to be inside it"""
clear_dir(TEMP_DIR)
os.chdir(TEMP_DIR)
def test_clear_temp_dir():
""" Clears temp dir"""
os.chdir(os.pardir)
clear_dir(TEMP_DIR, recreate=False)
def test_clear():
"""Clear test dir for tests"""
clear_dir(TEMP_DIR, recreate=False)