|
|
@ -17,6 +17,7 @@ from analytic_types.segment import Segment |
|
|
|
SHIFT_FACTOR = 0.05 |
|
|
|
SHIFT_FACTOR = 0.05 |
|
|
|
CONFIDENCE_FACTOR = 0.5 |
|
|
|
CONFIDENCE_FACTOR = 0.5 |
|
|
|
SMOOTHING_FACTOR = 5 |
|
|
|
SMOOTHING_FACTOR = 5 |
|
|
|
|
|
|
|
MEASUREMENT_ERROR = 0.05 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
def exponential_smoothing(series: pd.Series, alpha: float, last_smoothed_value: Optional[float] = None) -> pd.Series: |
|
|
|
def exponential_smoothing(series: pd.Series, alpha: float, last_smoothed_value: Optional[float] = None) -> pd.Series: |
|
|
@ -391,7 +392,7 @@ def find_parameters(segment_data: pd.Series, segment_from_index: int, pat_type: |
|
|
|
segment = flat_segment.dropna() |
|
|
|
segment = flat_segment.dropna() |
|
|
|
segment_median, segment_max_line, segment_min_line = utils.get_distribution_density(segment) |
|
|
|
segment_median, segment_max_line, segment_min_line = utils.get_distribution_density(segment) |
|
|
|
height = 0.95 * (segment_max_line - segment_min_line) |
|
|
|
height = 0.95 * (segment_max_line - segment_min_line) |
|
|
|
length = utils.find_length(segment_data, segment_min_line, segment_max_line, pat_type) |
|
|
|
length = utils.get_pattern_length(segment_data, segment_min_line, segment_max_line, pat_type) |
|
|
|
return height, length |
|
|
|
return height, length |
|
|
|
|
|
|
|
|
|
|
|
def find_pattern_center(segment_data: pd.Series, segment_from_index: int, pattern_type: str): |
|
|
|
def find_pattern_center(segment_data: pd.Series, segment_from_index: int, pattern_type: str): |
|
|
@ -404,14 +405,15 @@ def find_pattern_center(segment_data: pd.Series, segment_from_index: int, patter |
|
|
|
segment_cent_index = math.ceil((len(segment_data)) / 2) |
|
|
|
segment_cent_index = math.ceil((len(segment_data)) / 2) |
|
|
|
return segment_cent_index |
|
|
|
return segment_cent_index |
|
|
|
|
|
|
|
|
|
|
|
def find_length(segment_data: pd.Series, segment_min_line: float, segment_max_line: float, pat_type: str) -> int: |
|
|
|
def get_pattern_length(segment_data: pd.Series, segment_min_line: float, segment_max_line: float, pat_type: str) -> int: |
|
|
|
x_abscissa = np.arange(0, len(segment_data)) |
|
|
|
# TODO: move function to jump & drop merged model |
|
|
|
segment_max = max(segment_data) |
|
|
|
segment_max = max(segment_data) |
|
|
|
segment_min = min(segment_data) |
|
|
|
segment_min = min(segment_data) |
|
|
|
|
|
|
|
# TODO: use better way |
|
|
|
if segment_min_line <= segment_min: |
|
|
|
if segment_min_line <= segment_min: |
|
|
|
segment_min_line = segment_min * 1.05 |
|
|
|
segment_min_line = segment_min * (1 + MEASUREMENT_ERROR) |
|
|
|
if segment_max_line >= segment_max: |
|
|
|
if segment_max_line >= segment_max: |
|
|
|
segment_max_line = segment_max * 0.95 |
|
|
|
segment_max_line = segment_max * (1 - MEASUREMENT_ERROR) |
|
|
|
min_line = [] |
|
|
|
min_line = [] |
|
|
|
max_line = [] |
|
|
|
max_line = [] |
|
|
|
for i in range(len(segment_data)): |
|
|
|
for i in range(len(segment_data)): |
|
|
|