def get_experiment_choice(self) -> str:
""" Given a particular, many experiments would have been run. This method gets the choice
of experiment that the user wants to interact with
Returns
-------
str
The choice of experiment chosen by the user
"""
output_dirpath = pathlib.Path(OUTPUT_DIR)
experiment_dirnames = self.get_experiments_folder_names()
if len(experiment_dirnames) == 0:
self.msg_printer.fail(
f"There are no experiments for the model type {self.model_type_answer}"
)
exit(1)
experiment_choices = [
Choice(dirname) for dirname in experiment_dirnames
]
exp_choice = questionary.rawselect("Please select an experiment",
choices=experiment_choices,
qmark="❓").ask()
exp_choice_path = pathlib.Path(OUTPUT_DIR, exp_choice)
if not exp_choice_path.is_dir():
with self.msg_printer.loading(
f"Downloading experiment {exp_choice} from s3"):
self.s3util.download_folder(exp_choice,
download_only_best_checkpoint=True)
return str(output_dirpath.joinpath(exp_choice_path))
def pick_tasks(items, service_task):
# Allow users to select the task lists to move tasks from
todo_lists = []
for item in items:
todo_lists.append(item['title'])
selected_lists = (questionary.checkbox(
"Select List",
choices=todo_lists,
).ask() or ['My List'])
print(f"Moving tasks from {' and '.join(selected_lists)}.")
#Add a date range
min_date = questionary.text(
"Enter minimum date in format (yyyy-mm-dd)").ask()
min_date = min_date + 'T00:00:00.00Z'
print(min_date)
max_date_options = ['Now', 'Custom date']
selected_maxDate = (questionary.rawselect(
"Select maximum date of date range",
choices=max_date_options,
).ask() or "do nothing")
if selected_maxDate == 'Now':
max_date = datetime.now().strftime('%Y-%m-%dT%H:%M:%S.00Z')
else:
max_date = questionary.text(
"Enter maximum date in format (yyyy-mm-dd)").ask()
max_date = max_date + 'T00:00:00.00Z'
print(max_date)
#Filter completed tasks only
completed_tasks = dict()
for item in items:
if item['title'] in selected_lists:
task = service_task.tasks().list(tasklist=item['id'],
showHidden=1,
completedMin=min_date,
completedMax=max_date).execute()
# Filter tasks based on date range
for i in task['items']:
# We are concerned with date only, not time, hence splicing the date string to only include date
if str(i['updated'])[0:10] in completed_tasks:
completed_tasks[str(i['updated'])[0:10]].append(i['title'])
else:
completed_tasks[str(i['updated'])[0:10]] = [i['title']]
#Sorting the tasks in ascending order of updated date
completed_tasks = collections.OrderedDict(sorted(completed_tasks.items()))
#print(completed_tasks)
return completed_tasks
def ask_pystyle(**kwargs):
# create the question object
question = questionary.rawselect('What do you want to do?',
choices=[
'Order a pizza', 'Make a reservation',
Separator(), 'Ask opening hours',
'Talk to the receptionist'
],
style=custom_style_dope,
**kwargs)
# prompt the user for an answer
return question.ask()
def showMain(self):
questionary.text("What's your first name").ask()
questionary.password("What's your secret?").ask()
questionary.confirm("Are you amazed?").ask()
questionary.select(
"What do you want to do?",
choices=[
"Order a pizza", "Make a reservation", "Ask for opening hours"
],
).ask()
questionary.rawselect(
"What do you want to do?",
choices=[
"Order a pizza", "Make a reservation", "Ask for opening hours"
],
).ask()
questionary.checkbox("Select toppings", choices=["foo", "bar",
"bazz"]).ask()
questionary.path("Path to the projects version file").ask()
def save_qualifying_loans(qualifying_loans):
"""Saves the qualifying loans to a CSV file.
# this is a conditional
Args:
qualifying_loans (list of lists): The qualifying bank loans.
"""
# @TODO: Complete the usability dialog for savings the CSV Files.
# YOUR CODE HERE!
#name ="Michael" # string datatype
#number=24 # integer datatype
questionary.rawselect(
"Do you want to save the qualifying loans to the CSV file?",
choices=["Y,N"],).ask()
#answer = ""
#while answer not in ["y", "n"]:
# answer = raw_input("OK to push to continue [Y/N]? ").lower()
#save_csv_file = input("Do you want to save the qualifying loans to csv. Enter Y/N")
if save_csv_file == "y":
csvpath = raw_input("Enter the path to CSV file ? ").lower()
save_csv(csvpath,qualiying_loans)
else:
print("Not saving to csv")
def ask_model_type(self):
"""
Asks to chose a model amongst different model types that are available in sciwing
Returns
-------
str
Model type chosen by the user
"""
choices = self.return_model_type_choices()
model_type_question = questionary.rawselect(
"We have the following trained models. Chose one",
qmark="❓",
choices=choices,
)
return model_type_question.ask()
def ask_deletion() -> str:
""" Since this is deletion, we want confirmation, just to be sure
whether to keep the deleted folder locally or to remove it
Returns
-------
str
An yes or no answer to the question
"""
deletion_question = questionary.rawselect(
"Do you also want to delete the file locally. Caution! File will be removed locally",
qmark="❓",
choices=[Choice("yes"), Choice("no")],
)
deletion_answer = deletion_question.ask()
return deletion_answer
def ask_pystyle(**kwargs):
# create the question object
question = questionary.rawselect(
"What do you want to do?",
choices=[
"Order a pizza",
"Make a reservation",
Separator(),
"Ask opening hours",
"Talk to the receptionist",
],
style=custom_style_dope,
**kwargs,
)
# prompt the user for an answer
return question.ask()
def run_prompt(config: Config, session: Session, old_album: Album,
new_album: Album) -> Optional[Album]:
"""Runs the interactive prompt for the given album changes.
Args:
config: Moe config.
session: Current db session.
old_album: Original album to be added.
new_album: New album with all metadata changes. Will be compared against
``old_album``.
Returns:
The album to be added to the library.
"""
if old_album == new_album:
return old_album
existing_album = new_album.get_existing(session)
old_album.merge(existing_album, overwrite_album_info=False)
print(_fmt_album_changes(old_album, new_album)) # noqa: WPS421
prompt_choices: List[PromptChoice] = []
config.plugin_manager.hook.add_prompt_choice(prompt_choices=prompt_choices)
prompt_choices.sort(key=operator.attrgetter("shortcut_key"))
questionary_choices: List[questionary.Choice] = []
for prompt_choice in prompt_choices:
questionary_choices.append(
questionary.Choice(
title=prompt_choice.title,
shortcut_key=prompt_choice.shortcut_key,
value=prompt_choice.func,
))
prompt_choice_func = questionary.rawselect(
"What do you want to do?", choices=questionary_choices).ask()
if prompt_choice_func:
return prompt_choice_func(
config=config,
session=session,
old_album=old_album,
new_album=new_album,
)
return None
def ask_generate_report_or_interact():
""" Ask the user to chose between interacting with the model or generate the report
for all the experiments
Returns
-------
str
choce of the user as a string
"""
choices = [
Choice("Interact with model", "interact"),
Choice("Generate report (for all experiments)", "gen-report"),
]
generate_report_or_interact = questionary.rawselect(
"What would you like to do ", qmark="❓", choices=choices)
return generate_report_or_interact.ask()
import questionary
from examples import custom_style_dope
if __name__ == '__main__':
questionary.text("What's your first name").ask()
questionary.password("What's your secret?").ask()
questionary.confirm("Are you amazed?").ask()
questionary.select("What do you want to do?",
choices=[
'Order a pizza', 'Make a reservation',
'Ask for opening hours'
]).ask()
questionary.rawselect("What do you want to do?",
choices=[
'Order a pizza', 'Make a reservation',
'Ask for opening hours'
]).ask()
questionary.checkbox('Select toppings',
choices=["foo", "bar", 'bazz'],
style=custom_style_dope).ask()
def interact(self):
""" Interact with the user to explore different models
This method provides various options for exploration of the different models.
- ``See-Confusion-Matrix`` shows the confusion matrix on the test dataset.
- ``See-Examples-of-Classification`` is to explore correct and mis-classifications. You can provide two class numbers as in, ``2 3`` and it shows examples in the test dataset where text that belong to class ``2`` is classified as class ``3``.
- ``See-prf-table`` shows the precision recall and fmeasure per class.
- ``See-text`` - Manually enter text and look at the classification results.
Returns
-------
None
"""
exp_dir = self.get_experiment_choice()
exp_dir_path = pathlib.Path(exp_dir)
inference_func = self.model_type2inf_func[self.model_type_answer]
inference_client = inference_func(exp_dir)
inference_client.run_test()
while True:
choices = [
Choice("See-Confusion-Matrix"),
Choice("See-examples-of-Classifications"),
Choice("See-prf-table"),
Choice(title="Enter text ", value="enter_text"),
Choice("exit"),
]
if self.model_type_answer == "lstm-crf-scienceie-tagger":
choices.append(
Choice("official-results", "semeval_official_results"))
interaction_choice = questionary.rawselect(
"What would you like to do now", qmark="❓",
choices=choices).ask()
if interaction_choice == "See-Confusion-Matrix":
inference_client.print_confusion_matrix()
elif interaction_choice == "See-examples-of-Classifications":
misclassification_choice = questionary.text(
"Enter Two Classes separated by a space. [Hint: 1 2]").ask(
)
two_classes = [
int(class_) for class_ in misclassification_choice.split()
]
first_class, second_class = two_classes[0], two_classes[1]
sentences = inference_client.get_misclassified_sentences(
first_class, second_class)
self.msg_printer.divider(
f"Sentences with class {first_class} classified as {second_class}"
.capitalize())
for sentence in sentences:
print(sentence)
self.msg_printer.divider("")
elif interaction_choice == "See-prf-table":
inference_client.report_metrics()
elif interaction_choice == "enter_text":
text = questionary.text("Enter Text: ").ask()
tagged_string = inference_client.on_user_input(text)
print(tagged_string)
elif interaction_choice == "semeval_official_results":
dev_folder = pathlib.Path(SCIENCE_IE_DEV_FOLDER)
pred_folder = pathlib.Path(
REPORTS_DIR, f"science_ie_{exp_dir_path.stem}_results")
if not pred_folder.is_dir():
pred_folder.mkdir()
inference_client.generate_predict_folder(
dev_folder=dev_folder, pred_folder=pred_folder)
calculateMeasures(
folder_gold=str(dev_folder),
folder_pred=str(pred_folder),
remove_anno="rel",
)
elif interaction_choice == "exit":
self.msg_printer.text("See you again!")
exit(0)
import questionary
if __name__ == "__main__":
action = (questionary.rawselect(
"What do you want to do?",
choices=[
"Order a pizza", "Make a reservation", "Ask for opening hours"
],
).ask() or "do nothing")
print(f"Sorry, I can't {action}. Bye! 🙅")
def interact(self):
""" Interact with the user to explore different models
This method provides various options for exploration of the different models.
- ``See-Confusion-Matrix`` shows the confusion matrix on the test dataset.
- ``See-Examples-of-Classification`` is to explore correct and mis-classifications. You can provide two class numbers as in, ``2 3`` and it shows examples in the test dataset where text that belong to class ``2`` is classified as class ``3``.
- ``See-prf-table`` shows the precision recall and fmeasure per class.
- ``See-text`` - Manually enter text and look at the classification results.
"""
self.infer_obj.run_test()
while True:
choices = [
Choice("See-Confusion-Matrix"),
Choice("See-examples-of-Classifications"),
Choice("See-prf-table"),
Choice(title="Enter text ", value="enter_text"),
Choice(
title="If this is ScienceIE chose this to generate results",
value="science-ie-official-results",
),
Choice("exit"),
]
interaction_choice = rawselect("What would you like to do now",
qmark="❓",
choices=choices).ask()
if interaction_choice == "See-Confusion-Matrix":
self.infer_obj.print_confusion_matrix()
elif interaction_choice == "See-examples-of-Classifications":
misclassification_choice = ask_text(
"Enter Two Classes separated by a space. [Hint: 1 2]").ask(
)
two_classes = [
int(class_) for class_ in misclassification_choice.split()
]
first_class, second_class = two_classes[0], two_classes[1]
self.infer_obj.get_misclassified_sentences(
first_class, second_class)
elif interaction_choice == "See-prf-table":
self.infer_obj.report_metrics()
elif interaction_choice == "enter_text":
text = ask_text("Enter Text: ").ask()
tagged_string = self.infer_obj.on_user_input(text)
print(tagged_string)
elif interaction_choice == "semeval_official_results":
dev_folder = pathlib.Path(SCIENCE_IE_DEV_FOLDER)
pred_folder = ask_text(
message="Enter the directory path for storing results")
pred_folder = pathlib.Path(pred_folder)
if not pred_folder.is_dir():
pred_folder.mkdir()
self.infer_obj.generate_predict_folder(dev_folder=dev_folder,
pred_folder=pred_folder)
calculateMeasures(
folder_gold=str(dev_folder),
folder_pred=str(pred_folder),
remove_anno="rel",
)
elif interaction_choice == "exit":
self.msg_printer.text("See you again!")
exit(0)
