Skip to content

Commit

Permalink
black reformatting
Browse files Browse the repository at this point in the history
  • Loading branch information
@wflynny
wflynny committed May 28, 2020
1 parent b87c347 commit d7fa04c
Showing 1 changed file with 82 additions and 45 deletions.
127 changes: 82 additions & 45 deletions post_processing/hcs_data_processing.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -14,25 +14,44 @@ def parse_args() -> argparse.Namespace:
"Parse related files for use with the Single Cell Biology "
"Lab Opera Phenix High Content Screening platform"
),
prog=__file__
prog=__file__,
)
add_arg = lambda a, **v: parser.add_argument(a, **v)


add_arg(
"-i", dest="hcs_input_file", type=Path, required=True, help="Path to HCS input form"
"-i",
dest="hcs_input_file",
type=Path,
required=True,
help="Path to HCS input form",
)
add_arg(
"-r", dest="randomization_file", type=Path, required=True, help="Path to randomization CSV file"
"-r",
dest="randomization_file",
type=Path,
required=True,
help="Path to randomization CSV file",
)
add_arg(
"-s", dest="spectramax_file", type=Path, required=True, help="Path to Spectra Max export file"
"-s",
dest="spectramax_file",
type=Path,
required=True,
help="Path to Spectra Max export file",
)
add_arg(
"-o", dest="output_file", type=Path, required=True, help="Path to Spectra Max export file"
"-o",
dest="output_file",
type=Path,
required=True,
help="Path to Spectra Max export file",
)
add_arg(
"-p", dest="plot_pdf_file", type=Path, default=None, help="Path to Spectra Max export file"
"-p",
dest="plot_pdf_file",
type=Path,
default=None,
help="Path to Spectra Max export file",
)

return parser.parse_args()
Expand Down Expand Up @@ -108,21 +127,25 @@ def parse_phenix_metadata(path: Path) -> typing.Tuple[dict, pd.DataFrame]:

return description, plate_map


def assert_path_exists(name: str, path: Path) -> None:
if path is not None and not path.exists():
logger.error(f"{name.replace('_',' ').capitalize()}: [{path}] does not exist!")
print(path)
print(path.exists())
exit(2)


def assert_path_does_not_exist(name: str, path: Path) -> None:
if path is not None and path.exists():
logger.error(f"{name.replace('_',' ').capitalize()}: [{path}] already exist!")
exit(2)


def parse_description(excel_file: pd.ExcelFile) -> typing.Tuple[dict, pd.DataFrame]:
raw_description = excel_file.parse("Description", index_col=0, header=None).fillna("")
raw_description = excel_file.parse("Description", index_col=0, header=None).fillna(
""
)

description = raw_description.iloc[:18, :1].squeeze()
new_index = description.index.dropna()
Expand All @@ -139,11 +162,11 @@ def parse_description(excel_file: pd.ExcelFile) -> typing.Tuple[dict, pd.DataFra
return description, dispensing


def parse_daughter_plate_specs(excel_file: pd.ExcelFile) -> typing.Union[pd.DataFrame, None]:
def parse_daughter_plate_specs(
excel_file: pd.ExcelFile,
) -> typing.Union[pd.DataFrame, None]:
variables = excel_file.parse("Imaged Plates", index_col=0, header=0)
variable_names = variables[
"Variable Name"
]
variable_names = variables["Variable Name"]
valid_names = ~variable_names.isnull()
n_variables = np.sum(valid_names)
if n_variables == 0:
Expand All @@ -156,12 +179,12 @@ def parse_daughter_plate_specs(excel_file: pd.ExcelFile) -> typing.Union[pd.Data
return variables


def construct_dilution_plate(dispensing, dilution_constant=10**(1/2)):
dilutions = dispensing.iloc[1, :].values[:, None] * ((1 / dilution_constant) ** np.arange(12))
def construct_dilution_plate(dispensing, dilution_constant=10 ** (1 / 2)):
dilutions = dispensing.iloc[1, :].values[:, None] * (
(1 / dilution_constant) ** np.arange(12)
)
dilution_plate = pd.DataFrame(
dilutions,
index=list("ABCDEFGH"),
columns=pd.RangeIndex(1, 13)
dilutions, index=list("ABCDEFGH"), columns=pd.RangeIndex(1, 13)
)
dilution_plate = dilution_plate.stack().to_frame()
dilution_plate.index = dilution_plate.index.map("{0[0]}{0[1]}".format)
Expand All @@ -175,17 +198,22 @@ def parse_randomization_file(randomization_file):

sheets = []
for name in sheet_names:
sheet = pd.read_excel(
randomization_file, sheet_name=name, index_col=0
).dropna().iloc[:,0].values
sheet = (
pd.read_excel(randomization_file, sheet_name=name, index_col=0)
.dropna()
.iloc[:, 0]
.values
)
sheets.append(sheet.reshape(8, -1))

data_384 = np.block([sheets[:2], sheets[2:]])
df_384 = pd.DataFrame(
data_384,
index=list("ABCDEFGHIJKLMNOP"),
columns=pd.RangeIndex(1, 25)
).stack().to_frame()
df_384 = (
pd.DataFrame(
data_384, index=list("ABCDEFGHIJKLMNOP"), columns=pd.RangeIndex(1, 25)
)
.stack()
.to_frame()
)
df_384.index = df_384.index.map("{0[0]}{0[1]}".format)
df_384.columns = ["source_well_96"]
return df_384
Expand All @@ -197,14 +225,19 @@ def parse_spectramax_file(spectramax_file, encoding="utf-16"):
generated on Windows has carriage returns as line endings.
Due to this, the parsing is pretty fragile.
"""
data = pd.read_table(
spectramax_file,
encoding="utf-16",
engine="python",
sep="\t",
skiprows=2,
skipfooter=2
).iloc[:16, 2:26].stack().values
data = (
pd.read_table(
spectramax_file,
encoding="utf-16",
engine="python",
sep="\t",
skiprows=2,
skipfooter=2,
)
.iloc[:16, 2:26]
.stack()
.values
)
return data


Expand All @@ -219,8 +252,12 @@ def construct_dataframe(dilution, randomization_file, spectramax_file):

final["spectramax"] = parse_spectramax_file(spectramax_file)

final[["source_row_96", "source_col_96"]] = final.source_well_96.str.extract("([A-Z])(\d+)")
final[["row_384", "col_384"]] = final.index.to_series().str.extract("([A-Z])(\d+)", expand=True)
final[["source_row_96", "source_col_96"]] = final.source_well_96.str.extract(
"([A-Z])(\d+)"
)
final[["row_384", "col_384"]] = final.index.to_series().str.extract(
"([A-Z])(\d+)", expand=True
)

for col in final.columns:
final[col] = pd.to_numeric(final[col], errors="ignore")
Expand All @@ -238,7 +275,7 @@ def construct_dataframe(dilution, randomization_file, spectramax_file):

def construct_daughter_dataframe(final, daughter_info):
n_plates = daughter_info.shape[1]
expanded = pd.concat([final]*n_plates, axis=0)
expanded = pd.concat([final] * n_plates, axis=0)
repeat = lambda x: np.repeat(x, final.shape[0])
expanded["plate"] = repeat(np.arange(n_plates, dtype=int) + 1)
for key, row in daughter_info.iterrows():
Expand All @@ -261,24 +298,24 @@ def main(args: argparse.Namespace) -> None:

daughter_plate_info = parse_daughter_plate_specs(input_excel)

dilution_plate = construct_dilution_plate(dispensing, description["Serial dilution constant"])
dilution_plate = construct_dilution_plate(
dispensing, description["Serial dilution constant"]
)

dataframe = construct_dataframe(
dilution_plate,
args.randomization_file,
args.spectramax_file,
dilution_plate, args.randomization_file, args.spectramax_file,
)

daughter_dataframe = construct_daughter_dataframe(
dataframe,
daughter_plate_info
)
daughter_dataframe = construct_daughter_dataframe(dataframe, daughter_plate_info)

if args.plot_pdf_file is not None:
plot_plates()

dataframe.to_csv(args.output_file)
daughter_outfile = args.output_file.parent / f"{args.output_file.stem}-expanded{args.output_file.suffix}"
daughter_outfile = (
args.output_file.parent
/ f"{args.output_file.stem}-expanded{args.output_file.suffix}"
)
daughter_dataframe.to_csv(daughter_outfile)


Expand Down

0 comments on commit d7fa04c

Please sign in to comment.