Skip to content

Commit

Permalink
Revert "feat: SoftMax Pro - report non numeric values in error docume…
Browse files Browse the repository at this point in the history 
…nt (#713)"

This reverts commit 477d29e.
  • Loading branch information
@alejandro-salgado
alejandro-salgado committed Oct 16, 2024
1 parent bdb22b9 commit 81e24b8
Show file tree
Hide file tree
Showing 24 changed files with 25,395 additions and 89,737 deletions.
2 changes: 0 additions & 2 deletions src/allotropy/parsers/moldev_softmax_pro/softmax_pro_data_creator.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -121,8 +121,6 @@ def _create_measurements(plate_block: PlateBlock, position: str) -> list[Measure
total_measurement_time_setting=plate_block.header.read_time,
read_interval_setting=plate_block.header.read_interval,
number_of_scans_setting=plate_block.header.kinetic_points,
# Error documents
error_document=data_element.error_document,
)
for idx, data_element in enumerate(plate_block.iter_data_elements(position))
]
Expand Down
174 changes: 79 additions & 95 deletions src/allotropy/parsers/moldev_softmax_pro/softmax_pro_structure.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -10,15 +10,13 @@
import pandas as pd

from allotropy.allotrope.schema_mappers.adm.plate_reader.rec._2024._06.plate_reader import (
ErrorDocument,
MeasurementType,
ScanPositionSettingPlateReader,
)
from allotropy.exceptions import (
AllotropeConversionError,
get_key_or_error,
)
from allotropy.parsers.constants import NEGATIVE_ZERO
from allotropy.parsers.lines_reader import CsvReader
from allotropy.parsers.utils.pandas import rm_df_columns, SeriesData, set_columns
from allotropy.parsers.utils.uuids import random_uuid_str
Expand Down Expand Up @@ -114,15 +112,13 @@ class GroupDataElement:
position: str
plate: str
entries: list[GroupDataElementEntry]
errors: list[ErrorDocument]


@dataclass(frozen=True)
class GroupSampleData:
identifier: str
data_elements: list[GroupDataElement]
aggregated_entries: list[GroupDataElementEntry]
aggregated_errors: list[ErrorDocument]

@classmethod
def create(cls, data: pd.DataFrame) -> GroupSampleData:
Expand All @@ -131,7 +127,7 @@ def create(cls, data: pd.DataFrame) -> GroupSampleData:
identifier = top_row[str, "Sample"]
data = rm_df_columns(data, r"^Sample$|^Standard Value|^R$|^Unnamed: \d+$")
# Columns are considered "numeric" if the value of the first row is a float
# Non-numeric values such as "Mask" and "Range?" will be reported as errors.
# "Mask" and "Range?" are special cases that will be considered NaN.
numeric_columns = [
column
for column in data.columns
Expand All @@ -147,43 +143,40 @@ def create(cls, data: pd.DataFrame) -> GroupSampleData:
else:
normal_columns.append(column)

data_elements = []
for row in row_data:
entries, errors = cls._get_entries_and_errors(row, normal_columns)
data_elements.append(
return GroupSampleData(
identifier=identifier,
data_elements=[
GroupDataElement(
sample=identifier,
position=row[str, ["Well", "Wells"]],
plate=row[str, "WellPlateName"],
entries=entries,
errors=errors,
entries=[
element_entry
for column_name in normal_columns
if (element_entry := cls._get_element_entry(row, column_name))
is not None
],
)
)

aggregated_entries, aggregated_errors = cls._get_entries_and_errors(
top_row, aggregated_columns
)

return GroupSampleData(
identifier=identifier,
data_elements=data_elements,
aggregated_entries=aggregated_entries,
aggregated_errors=aggregated_errors,
for row in row_data
],
aggregated_entries=[
element_entry
for column_name in aggregated_columns
if (element_entry := cls._get_element_entry(top_row, column_name))
is not None
],
)

@classmethod
def _get_entries_and_errors(
cls, data_row: SeriesData, column_names: list[str]
) -> tuple[list[GroupDataElementEntry], list[ErrorDocument]]:
entries = []
errors = []
for column in column_names:
value = data_row.get(float, column)
if value is not None:
entries.append(GroupDataElementEntry(column, value))
elif (error := data_row.get(str, column)) is not None:
errors.append(ErrorDocument(error, column))
return entries, errors
def _get_element_entry(
cls, data_row: SeriesData, column_name: str
) -> GroupDataElementEntry | None:
if (value := data_row.get(float, column_name)) is not None:
return GroupDataElementEntry(
name=column_name,
value=value,
)
return None


@dataclass(frozen=True)
Expand Down Expand Up @@ -311,7 +304,6 @@ class DataElement:
wavelength: float
position: str
value: float
error_document: list[ErrorDocument]
elapsed_time: list[float] = field(default_factory=list)
kinetic_measures: list[float | None] = field(default_factory=list)
sample_id: str | None = None
Expand All @@ -334,58 +326,54 @@ class PlateWavelengthData:

@staticmethod
def create(
header: PlateHeader,
plate_name: str,
temperature: float | None,
elapsed_time: float | None,
wavelength: float,
df_data: pd.DataFrame,
) -> PlateWavelengthData:
# Since value is required for the measurement class (absorbance, luminescense and fluorescence)
# we don't store data for NaN values
# TODO: Report error documents for NaN values
data = {
f"{num_to_chars(row_idx)}{col}": raw_value
f"{num_to_chars(row_idx)}{col}": value
for row_idx, *row_data in df_data.itertuples()
for col, raw_value in zip(df_data.columns, row_data, strict=True)
if (value := try_non_nan_float_or_none(raw_value)) is not None
}

data_elements = {}
for position, raw_value in data.items():
value = try_non_nan_float_or_none(raw_value)
if value is None and elapsed_time is not None:
msg = f"Missing kinetic measurement for well position {position} at {elapsed_time}s."
raise AllotropeConversionError(msg)

data_elements[str(position)] = DataElement(
uuid=random_uuid_str(),
plate=header.name,
temperature=temperature,
wavelength=wavelength,
position=str(position),
value=NEGATIVE_ZERO if value is None else value,
error_document=(
[ErrorDocument(str(raw_value), header.read_mode)]
if value is None
else []
),
elapsed_time=[elapsed_time] if elapsed_time is not None else [],
kinetic_measures=[value] if elapsed_time is not None else [],
)

return PlateWavelengthData(wavelength, data_elements)
return PlateWavelengthData(
wavelength,
data_elements={
str(position): DataElement(
uuid=random_uuid_str(),
plate=plate_name,
temperature=temperature,
wavelength=wavelength,
position=str(position),
value=value,
elapsed_time=[elapsed_time] if elapsed_time is not None else [],
kinetic_measures=[value] if elapsed_time is not None else [],
)
for position, value in data.items()
},
)

def update_kinetic_data_elements(
self, elapsed_time: float, df_data: pd.DataFrame
) -> None:
data = {
f"{num_to_chars(row_idx)}{col}": try_non_nan_float_or_none(raw_value)
f"{num_to_chars(row_idx)}{col}": value
for row_idx, *row_data in df_data.itertuples()
for col, raw_value in zip(df_data.columns, row_data, strict=True)
if (value := try_non_nan_float_or_none(raw_value)) is not None
}
for position, value in data.items():
if value is None:
try:
self.data_elements[position].elapsed_time.append(elapsed_time)
self.data_elements[position].kinetic_measures.append(value)
except KeyError as e:
msg = f"Missing kinetic measurement for well position {position} at {elapsed_time}s."
raise AllotropeConversionError(msg)

self.data_elements[position].elapsed_time.append(elapsed_time)
self.data_elements[position].kinetic_measures.append(value)
raise AllotropeConversionError(msg) from e


@dataclass(frozen=True)
Expand Down Expand Up @@ -467,7 +455,7 @@ def _get_wavelength_data(
end = start + header.num_columns
wavelength_data.append(
PlateWavelengthData.create(
header=header,
plate_name=header.name,
temperature=temperature,
elapsed_time=elapsed_time,
wavelength=header.wavelengths[idx],
Expand Down Expand Up @@ -536,6 +524,8 @@ def create(

def iter_data_elements(self, position: str) -> Iterator[DataElement]:
for wavelength_data in self.raw_data.wavelength_data:
if position not in wavelength_data.data_elements:
continue
yield wavelength_data.data_elements[position]


Expand All @@ -545,29 +535,26 @@ class TimeMeasurementData:

@staticmethod
def create(
header: PlateHeader,
plate_name: str,
wavelength: float,
row: pd.Series[float],
) -> TimeMeasurementData:
temperature = try_non_nan_float_or_none(str(row.iloc[1]))
data_elements = {}

for position, raw_value in row.iloc[2:].items():
value = try_non_nan_float_or_none(raw_value)
error_document = []
if value is None:
error_document.append(ErrorDocument(str(raw_value), header.read_mode))
data_elements[str(position)] = DataElement(
uuid=random_uuid_str(),
plate=header.name,
temperature=temperature,
wavelength=wavelength,
position=str(position),
value=NEGATIVE_ZERO if value is None else value,
error_document=error_document,
)

return TimeMeasurementData(data_elements)
return TimeMeasurementData(
data_elements={
str(position): DataElement(
uuid=random_uuid_str(),
plate=plate_name,
temperature=temperature,
wavelength=wavelength,
position=str(position),
value=value,
)
for position, raw_value in row.iloc[2:].items()
if (value := try_non_nan_float_or_none(str(raw_value))) is not None
},
)


@dataclass(frozen=True)
Expand All @@ -578,7 +565,7 @@ class TimeWavelengthData:
@staticmethod
def create(
reader: CsvReader,
header: PlateHeader,
plate_name: str,
wavelength: float,
columns: pd.Series[str],
) -> TimeWavelengthData:
Expand All @@ -590,7 +577,7 @@ def create(
return TimeWavelengthData(
wavelength=wavelength,
measurement_data=[
TimeMeasurementData.create(header, wavelength, row)
TimeMeasurementData.create(plate_name, wavelength, row)
for _, row in data.iterrows()
],
)
Expand All @@ -611,7 +598,7 @@ def create(reader: CsvReader, header: PlateHeader) -> TimeRawData:
wavelength_data=[
TimeWavelengthData.create(
reader,
header,
header.name,
wavelength,
columns,
)
Expand Down Expand Up @@ -671,6 +658,8 @@ def create(
def iter_data_elements(self, position: str) -> Iterator[DataElement]:
for wavelength_data in self.raw_data.wavelength_data:
for measurement_data in wavelength_data.measurement_data:
if position not in measurement_data.data_elements:
continue
yield measurement_data.data_elements[position]


Expand Down Expand Up @@ -1094,7 +1083,6 @@ class StructureData:
def create(reader: CsvReader) -> StructureData:
block_list = BlockList.create(reader)

# update sample_id if it was reported in the group blocks and include erros from calculated data
for group_block in block_list.group_blocks:
for group_sample_data in group_block.group_data.sample_data:
for group_data_element in group_sample_data.data_elements:
Expand All @@ -1103,9 +1091,5 @@ def create(reader: CsvReader) -> StructureData:
group_data_element.position
):
data_element.sample_id = group_data_element.sample
data_element.error_document += (
group_data_element.errors
+ group_sample_data.aggregated_errors
)

return StructureData(block_list)
Loading

0 comments on commit 81e24b8

Please sign in to comment.