การแบ่งข้อมูลตามเวลา

def train_test_split_date(df, split_col, test_days=____):
  """Calculate the date to split test and training sets"""
  # Find how many days our data spans
  max_date = df.____({____: ____}).collect()[0][0]
  min_date = df.____({____: ____}).collect()[0][0]
  # Subtract an integer number of days from the last date in dataset
  split_date = ____ - timedelta(days=____)
  return split_date

# Find the date to use in spitting test and train
split_date = train_test_split_date(df, ____)

# Create Sequential Test and Training Sets
____ = df.where(df[____] < split_date) 
____ = df.where(df[____] >= split_date).where(df[____] <= split_date)