Skip to content

Commit

Permalink
Merge pull request #58 from MetaSys-LISBP/dev_config_data_selector
Browse files Browse the repository at this point in the history 
Dev config data selector
  • Loading branch information
@llegregam
llegregam committed May 21, 2024
2 parents 7cd6781 + 382a635 commit 17c6d09
Show file tree
Hide file tree
Showing 3 changed files with 173 additions and 126 deletions.
4 changes: 3 additions & 1 deletion physiofit/__main__.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -21,6 +21,7 @@ def get_last_version():
except Exception:
pass


def main():
"""The main routine"""

Expand All @@ -33,5 +34,6 @@ def main():
path_to_app = path_to_app / "ui/gui.py"
run(["streamlit", "run", str(path_to_app)])


if __name__ == "__main__":
sys.exit(main())
sys.exit(main())
3 changes: 2 additions & 1 deletion physiofit/base/io.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -648,7 +648,8 @@ def check_data_path(self):
"""
if self.path_to_data:
return Path(self.path_to_data).is_file()
return None
else:
return False

def update_model(self, model):

Expand Down
292 changes: 168 additions & 124 deletions physiofit/ui/gui.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -71,17 +71,23 @@ def check_uptodate(self):
except Exception:
pass

def _build_flux_menu(self):
"""Build the starting menu with the data upload button"""
@staticmethod
def _get_data_path_config_context():

self.data_file = st.file_uploader(
"Load a data file or a json configuration file",
key="data_uploader",
accept_multiple_files=False
)
# Set up tkinter for directory chooser
root = tk.Tk()
root.withdraw()

if self.data_file:
self._initialize_opt_menu()
# Make folder picker dialog appear on top of other windows
root.wm_attributes('-topmost', 1)

return str(Path(
filedialog.askopenfilename(
master=root,
title="Select the data file",
filetypes=[("Data files", "*.tsv *.txt")]
)
))

def _initialize_opt_menu(self):
"""
Expand All @@ -97,20 +103,37 @@ def _initialize_opt_menu(self):
self.config_parser = self.io.read_yaml(self.data_file)
# Check if the data path exists, if not open up prompt to
# select file
# if not self.config_parser.check_data_path():
# self._output_directory_selector()
# else:
if hasattr(st.session_state, "config_parser_data_path"):
# If the path is already in session state, use it
self.config_parser.path_to_data = st.session_state[
"config_parser_data_path"]
if not self.config_parser.check_data_path():
st.info(
"File path in configuration file is incorrect or "
"missing. Loading file from prompt."
)
# Send path to session state to avoid multiple prompts
st.session_state["config_parser_data_path"] = (
self._get_data_path_config_context())
self.config_parser.path_to_data = st.session_state[
"config_parser_data_path"
]
st.info(f"Input data: {self.config_parser.path_to_data}")
# Load data into io_handler
self.io.data = self.io.read_data(
self.config_parser.path_to_data)
str(self.config_parser.path_to_data)
)
input_loaded = True
except Exception:
st.error(
"An error has occurred when reading the yaml "
"configuration file.")
"configuration file and/or loading the input data."
)
raise
elif file_extension in ["tsv", "txt"]:
try:
self.io.data = self.io.read_data(self.data_file)
input_loaded = True
# Initialize default SDs
self.defaults["sd"].update({
"X": 0.5
Expand All @@ -127,119 +150,136 @@ def _initialize_opt_menu(self):
f"files or json for configuration files. Detected file: "
f"{self.data_file.name}")

if "experiments" not in self.io.data.columns:
raise ValueError(
"'experiments' column missing from dataset"
if input_loaded:
# Reset config_parser path to data if it exists to handle new
# data file paths
if hasattr(st.session_state, "config_parser_data_path"):
del st.session_state["config_parser_data_path"]
if "experiments" not in self.io.data.columns:
raise ValueError(
"'experiments' column missing from dataset"
)
self.io.data = self.io.data.sort_values(
["experiments", "time"], ignore_index=True
)
self.io.data = self.io.data.sort_values(
["experiments", "time"], ignore_index=True
)

try:
# Initialize the list of available models
self.io.get_models()
except Exception:
st.error(
f"An error has occurred when listing models from the models "
f"folder: \n{Path(__file__).parent / 'models'}. Please correct"
f" the model or submit an issue at "
f"github.com/MetaSys-LISBP/PhysioFit/issues")
raise

# Build menu
submitted = self._initialize_opt_menu_widgets(
file_extension
)

if submitted:
handler = logging.FileHandler(self.io.res_path / "log.txt", "w")
stream = logging.StreamHandler()
handler.setLevel(logging.INFO)
stream.setLevel(logging.INFO)
if self.debug_mode:
handler.setLevel(logging.DEBUG)
stream.setLevel(logging.DEBUG)
formatter = logging.Formatter('%(asctime)s - %(name)s - %('
'levelname)s - %(message)s')
handler.setFormatter(formatter)
logger.addHandler(handler)
logger.addHandler(stream)
try:
self._get_data_from_session_state()
# Initialize the list of available models
self.io.get_models()
except Exception:
st.error("An error has occurred when initializing the model")
st.error(
f"An error has occurred when listing models from the "
f"models folder: \n{Path(__file__).parent / 'models'}. "
f"Please correct the model or submit an issue at "
f"github.com/MetaSys-LISBP/PhysioFit/issues")
raise
if not self.io.wkdir:
raise ValueError("No output directory selected")
self.config_parser = ConfigParser(
path_to_data=self.io.wkdir / self.data_file.name,
selected_model=self.model,
sds=self.sd,
mc=self.mc,
iterations=self.iterations
)

full_dataframe = self.io.data.copy()
results_path = copy(self.io.res_path)
experiments = list(self.io.data["experiments"].unique())
self.io.multiple_experiments = []
for experiment in experiments:
logger.info(f"Running optimization for {experiment}")
with st.spinner(f"Running optimization for {experiment}"):
# final_table_dict = {}
self.model.data = full_dataframe[
full_dataframe["experiments"] == experiment
].drop("experiments", axis=1).copy()

self.io.res_path = results_path / str(experiment)
if not self.io.res_path.is_dir():
self.io.res_path.mkdir(parents=True)
# Initialize the fitter object
self.io.names = self.io.data.columns[1:].to_list()
kwargs = self._build_fitter_kwargs()
logger.info("Run options for the fitter:")
for key, value in kwargs.items():
logger.info(f"{key} : {value}")
fitter = self.io.initialize_fitter(
self.model.data,
model=kwargs["model"],
mc=kwargs["mc"],
iterations=kwargs["iterations"],
sd=kwargs["sd"],
debug_mode=kwargs["debug_mode"]
)
# Do the work
fitter.optimize()
if self.mc:
fitter.monte_carlo_analysis()
fitter.khi2_test()
df = pd.DataFrame.from_dict(
fitter.parameter_stats,
orient="columns"
)
df.index = [
f"{experiment} {param}" for param in
fitter.model.parameters.keys()
]
st.write(df)
logger.info(f"Results for {experiment}: \n{df}")
self.io.multiple_experiments.append(df)

# Export results
self.io.output_report(fitter, self.io.res_path)
self.io.plot_data(fitter)
self.io.output_plots(fitter, self.io.res_path)
with st.expander(f"{experiment} plots"):
for fig in self.io.figures:
st.pyplot(fig[1])
self.io.output_pdf(fitter, self.io.res_path)
# Reset figures to free memory
self.io.figures = []
self.config_parser.export_config(self.io.res_path)
self.io.data = full_dataframe
logger.info(f"Resulting dataframe: \n{full_dataframe}")
self.io.res_path = results_path
self.io.output_recap(results_path)
# Build menu
submitted = self._initialize_opt_menu_widgets(file_extension)

if submitted:
handler = logging.FileHandler(self.io.res_path / "log.txt",
"w")
stream = logging.StreamHandler()
handler.setLevel(logging.INFO)
stream.setLevel(logging.INFO)
if self.debug_mode:
handler.setLevel(logging.DEBUG)
stream.setLevel(logging.DEBUG)
formatter = logging.Formatter('%(asctime)s - %(name)s - %('
'levelname)s - %(message)s')
handler.setFormatter(formatter)
logger.addHandler(handler)
logger.addHandler(stream)
try:
self._get_data_from_session_state()
except Exception:
st.error(
"An error has occurred when initializing the model")
raise
if not self.io.wkdir:
raise ValueError("No output directory selected")
self.config_parser = ConfigParser(
path_to_data=self.io.wkdir / self.data_file.name,
selected_model=self.model,
sds=self.sd,
mc=self.mc,
iterations=self.iterations
)

full_dataframe = self.io.data.copy()
results_path = copy(self.io.res_path)
experiments = list(self.io.data["experiments"].unique())
self.io.multiple_experiments = []
for experiment in experiments:
logger.info(f"Running optimization for {experiment}")
with st.spinner(f"Running optimization for {experiment}"):
# final_table_dict = {}
self.model.data = full_dataframe[
full_dataframe["experiments"] == experiment
].drop("experiments", axis=1).copy()

self.io.res_path = results_path / str(experiment)
if not self.io.res_path.is_dir():
self.io.res_path.mkdir(parents=True)
# Initialize the fitter object
self.io.names = self.io.data.columns[1:].to_list()
kwargs = self._build_fitter_kwargs()
logger.info("Run options for the fitter:")
for key, value in kwargs.items():
logger.info(f"{key} : {value}")
fitter = self.io.initialize_fitter(
self.model.data,
model=kwargs["model"],
mc=kwargs["mc"],
iterations=kwargs["iterations"],
sd=kwargs["sd"],
debug_mode=kwargs["debug_mode"]
)
# Do the work
fitter.optimize()
if self.mc:
fitter.monte_carlo_analysis()
fitter.khi2_test()
df = pd.DataFrame.from_dict(
fitter.parameter_stats,
orient="columns"
)
df.index = [
f"{experiment} {param}" for param in
fitter.model.parameters.keys()
]
st.write(df)
logger.info(f"Results for {experiment}: \n{df}")
self.io.multiple_experiments.append(df)

# Export results
self.io.output_report(fitter, self.io.res_path)
self.io.plot_data(fitter)
self.io.output_plots(fitter, self.io.res_path)
with st.expander(f"{experiment} plots"):
for fig in self.io.figures:
st.pyplot(fig[1])
self.io.output_pdf(fitter, self.io.res_path)
# Reset figures to free memory
self.io.figures = []
self.config_parser.export_config(self.io.res_path)
self.io.data = full_dataframe
logger.info(f"Resulting dataframe: \n{full_dataframe}")
self.io.res_path = results_path
self.io.output_recap(results_path)

def _build_flux_menu(self):
"""Build the starting menu with the data upload button"""

self.data_file = st.file_uploader(
"Load a data file or a json configuration file",
key="data_uploader",
accept_multiple_files=False
)

if self.data_file:
self._initialize_opt_menu()

def silent_sim(self):

Expand Down Expand Up @@ -333,6 +373,10 @@ def _initialize_opt_menu_widgets(self, file_extension):
self.config_parser.path_to_data).resolve().parent
self.io.res_path = self.io.wkdir / (
self.io.wkdir.name + "_res")
st.info(f"Data loaded from path: "
f"{self.config_parser.path_to_data}. ")
st.info(f"Output directory set to the same directory as "
f"the data file: {self.io.res_path}")

# Build the form for advanced parameters
form = st.form("Parameter_form")
Expand Down Expand Up @@ -404,12 +448,12 @@ def _initialize_opt_menu_widgets(self, file_extension):
with col2:
st.write("Parameter Value")
for key, value in self.model.args[
param].items():
param].items():
st.text_input(
label="label", # Unused
label_visibility="collapsed",
value=value if self.config_parser is
None else
None else
self.config_parser.model["args"][key],
key=f"Fixed_{param}_value_{key}"
)
Expand Down Expand Up @@ -507,7 +551,7 @@ def _get_data_from_session_state(self):
for key in self.model.args[param].keys():
try:
if st.session_state[
f"Fixed_{param}_value_{key}"] == "0":
f"Fixed_{param}_value_{key}"] == "0":
self.model.args[param][key] = 0
else:
self.model.args[param][key] = literal_eval(
Expand Down

0 comments on commit 17c6d09

Please sign in to comment.