def sort_data(dataset):
print("Sorting data into database...")
sort_data_time_start = time.time()
sql_connection = sqlite3.connect(database.db_name)
cur = sql_connection.cursor()
# Lets not f**k up sorted data by creating duplicate data
if cur.execute('select count(*) from data;').fetchall()[0][0] != 0:
print("There is already rows in the table...")
return
for set_num, data in enumerate(dataset):
print("Sorting set",set_num)
sort_data_set_time_start = time.time()
for section_number, section in enumerate(data.get_values()):
try:
threshold = jmath.get_peak_threshold(section)
except ZeroDivisionError:
# Skip section if stddev is not possible
break
for pos, value in enumerate(section):
classification = 'peak' if value > threshold else 'nopeak'
cur.execute("insert into data ('value','vector','class') values (?,?,?)",[value, json.dumps(get_vector(section, pos, jmath.vector_length)), classification])
sql_connection.commit()
print("Sorting set",set_num,"took",time.time()-sort_data_set_time_start,"seconds...")
sql_connection.close()
print("Sorting all sets took",time.time()-sort_data_time_start,"seconds...")
def get_classification(numbers, value): if value > jmath.get_peak_threshold(numbers): return "peak" return "nopeak"
