def slot_convert(self, kind, column_idx):
column = self.dataset.get_column_name(column_idx)
# Determine which function to run in order to convert
if kind == "binary":
convert_func = clarite.modify.make_binary
elif kind == "categorical":
convert_func = clarite.modify.make_categorical
elif kind == "continuous":
convert_func = clarite.modify.make_continuous
else:
show_warning(
"Error converting to unknown",
"Converting to unknown type is not yet implemented.",
)
return
# Try to convert, reporting any errors
try:
self.appctx.update_data(convert_func(self.dataset.df, only=column))
except ValueError as e:
show_warning(f"Error converting to {kind}", str(e))
return
# Log successful conversion
info_str = f"\nConverted '{column}' to {kind} in '{self.dataset.get_selector_name()}'\n"
self.appctx.log_info("\n" + "=" * 80 + info_str + "=" * 80)
self.appctx.log_python(f"{self.dataset.get_python_name()} = "
f"clarite.modify.make_{kind}("
f"data={self.dataset.get_python_name()}, "
f"skip=[], "
f"only=['{column}'])")
def submit(self):
# TODO: Add any warnings here
if self.new_name == self.old_name:
self.reject()
elif self.new_name in list(self.data):
show_warning(
"Error renaming column",
f"The specified column ('{self.new_name}') already exists",
)
else:
self.accept()
def submit(self):
if self.data_name is not None and self.data_name in [
d.name for d in self.appctx.datasets
]:
show_warning(
"Dataset already exists",
f"A dataset named '{self.data_name}' already exists.\n"
f"Use a different name or clear the dataset name field.",
)
else:
print(f"Calculating Percent NA")
# Run with a progress dialog
RunProgress.run_with_progress(
progress_str="Calculating Percent NA...",
function=self.get_func(),
slot=self.appctx.add_dataset,
parent=self,
)
self.log_command()
self.accept()
def submit(self):
if self.data_name is not None and self.data_name in [
d.name for d in self.appctx.datasets
]:
show_warning(
"Dataset already exists",
f"A dataset named '{self.data_name}' already exists.\n"
f"Use a different name or clear the dataset name field.",
)
elif "converged" not in list(self.dataset.df):
show_warning("Incorrect Data Input",
"A 'converged' column must be present")
elif "pvalue" not in list(self.dataset.df):
show_warning("Incorrect Data Input",
"A 'pvalue' column must be present")
else:
print(f"Adding corrected P-values...")
# Run with a progress dialog
if self.data_name is None:
slot = self.appctx.update_data
else:
slot = self.appctx.add_dataset
RunProgress.run_with_progress(
progress_str="Adding corrected P-values...",
function=self.get_func(),
slot=slot,
parent=self,
)
self.log_command()
self.accept()
def launch_get_category_file(self):
"""Launch a dialog to load a file which specified categories for each variable"""
options = QFileDialog.Options()
options |= QFileDialog.DontUseNativeDialog
filename, _ = QFileDialog.getOpenFileName(
self,
f"Load Variable Category File",
"",
f"TSV Files (*.tsv *.txt)",
options=options,
)
# Set filename
if len(filename) == 0:
return
# Read file
try:
categories = pd.read_csv(filename, sep="\t")
except Exception as e:
show_warning("Variable Categories File Error",
f"Error reading file: {str(e)}")
return
# Must have two columns
if len(list(categories)) != 2:
show_warning(
"Variable Categories File Error",
f"Expected 2 columns, found {len(list(categories)):,} columns",
)
return
# Set columns and convert to a dictionary
categories.columns = ["variable", "category"]
categories = categories.set_index("variable")
categories = categories.to_dict()["category"]
self.categories = categories
self.category_file_btn.setText(
f"{len(categories):,} variable categories loaded")
def submit(self):
if self.data_name is not None and self.data_name in [
d.name for d in self.appctx.datasets
]:
show_warning(
"Dataset already exists",
f"A dataset named '{self.data_name}' already exists.\n"
f"Use a different name or clear the dataset name field.",
)
elif self.outcome is None:
show_warning("Missing Parameter", "A phenotype must be selected")
else:
print(f"Running EWAS...")
# Run with a progress dialog
RunProgress.run_with_progress(
progress_str="Running EWAS...",
function=self.get_func(),
slot=self.appctx.add_dataset,
parent=self,
)
self.log_command()
self.accept()
def slot_rename(self, column_idx):
column = self.dataset.get_column_name(column_idx)
# Show a dialog
new_name = RenameDialog.get_new_name(
data=self.appctx.datasets[self.appctx.current_dataset_idx].df,
column=column,
parent=self.parent(),
)
if new_name is None:
return
try:
self.appctx.update_data(
self.dataset.df.rename(columns={column: new_name}))
except ValueError as e:
show_warning("Error renaming column", str(e))
return
# Log successful rename
info_str = f"\nRenamed '{column}' to '{new_name}' in '{self.dataset.get_selector_name()}'\n"
self.appctx.log_info("\n" + "=" * 80 + info_str + "=" * 80)
self.appctx.log_python(f"{self.dataset.get_python_name()} = "
f"{self.dataset.get_python_name()}.rename("
f"columns={repr({column: new_name})})")
def launch_get_weight_specific(self):
"""Launch a dialog to load a file matching variables to weights"""
options = QFileDialog.Options()
options |= QFileDialog.DontUseNativeDialog
filename, _ = QFileDialog.getOpenFileName(
self,
f"Load Specific Weights File",
"",
f"TSV Files (*.tsv *.txt)",
options=options,
)
# Set filename
if len(filename) == 0:
return
# Read file
try:
weights = pd.read_csv(filename, sep="\t")
except Exception as e:
show_warning("Specific Weights File Error",
f"Error reading file: {str(e)}")
return
# Must have two columns
if len(list(weights)) != 2:
show_warning(
"Specific Weights File Error",
f"Expected 2 columns, found {len(list(weights)):,} columns",
)
return
# Set columns and convert to a dictionary
weights.columns = ["variable", "weight"]
weights = weights.set_index("variable")
weights = weights.to_dict()["weight"]
# Check that some variables/weights matched
unique_vars = len(set(weights.keys()) & set(list(self.dataset.df)))
unique_weights = len(
set(weights.values()) & set(list(self.survey_df.df)))
missing_weights = (set(list(self.dataset.df)) - set(weights.keys()) -
set(weights.values()) - set(self.covariates) -
{self.outcome, self.cluster, self.strata, self.fpc})
if unique_vars < 1:
show_warning(
"Specific Weights File Error",
f"Loaded {filename}\n"
"No variables matched columns in the input data.\n\n"
"The first column of the specific weights file must list variable names and "
"a header line must be present.",
)
elif unique_weights < 1:
show_warning(
"Specific Weights File Error",
f"Loaded {filename}\n"
"No weights matched columns in the survey data.\n\n"
"The second column of the specific weights file must list weight names and "
"a header line must be present.",
)
elif len(missing_weights) > 0 and len(missing_weights) <= 5:
show_warning(
"Specific Weights File Error",
f"Loaded {filename}\n"
"Some variables are missing weights:\n\n"
f"{', '.join(sorted(list(missing_weights)))}",
)
elif len(missing_weights) > 5:
show_warning(
"Specific Weights File Error",
f"Loaded {filename}\n"
"More than 5 variables are missing weights, including:\n\n"
f"{', '.join(sorted(list(missing_weights))[:5])}",
)
else:
self.weights = weights
self.weight_specific_btn.setText(
f"{unique_weights:,} different weights "
f"assigned to {unique_vars:,} variables")