from otree.api import ( models, widgets, BaseConstants, BaseSubsession, BaseGroup, BasePlayer, Currency as c, currency_range, ) from enum import Enum import random import json author = 'Josie_NTULAB' doc = """ 決策實驗-問卷部分 """ class WaitingPeriod(object): list = [] class GainedAmount(object): list = [] today = 100 class Treatment(object): participant_ids = [] def prepare_participant_ids_if_needed(all_players): noNeedToReCreate = len(Treatment.participant_ids) == len(all_players) if noNeedToReCreate: return Treatment.participant_ids.clear() for p in all_players: pid = p.participant.id_in_session if pid not in Treatment.participant_ids: Treatment.participant_ids.append(pid) available_pronoun_list = ['我'] pronoun_list = [] def prepare_pronoun_list(all_players): # 事先準備好所有受試者的 pronoun 們 Treatment.pronoun_list.clear() Treatment.prepare_participant_ids_if_needed(all_players) for each_pronoun in Treatment.available_pronoun_list: count = int(len(Treatment.participant_ids) / len(Treatment.available_pronoun_list)) # 例如有5人，平均分配兩組 treatment，那就是 5/2 = 2（不取餘數） Treatment.pronoun_list.extend( [each_pronoun] * count ) # “python - Create list of single item repeated N times - Stack Overflow” https://stackoverflow.com/questions/3459098/create-list-of-single-item-repeated-n-times while len(Treatment.pronoun_list) < len(Treatment.participant_ids): # 若因前面 for 迴圈中不能整除，也就是 `available_pronoun_list` 不能完平均地分配給所有 participants，則最後會不夠，在此補上。（如果 `available_pronoun_list` 就兩個，那最多就差1，也就是最多會差到(count - 1)個） Treatment.pronoun_list.append(random.choice(Treatment.available_pronoun_list)) # 剩下的就隨機挑選 random.shuffle(Treatment.pronoun_list) # 把例如 [你, 你, 你, 受試者, 受試者] 打亂成 [受試者, 你, 你, 受試者, 你] def get_pronoun(player): participant = player.participant # lazy loading: 若不存在就取得並存起來，若已存在就直接取出 if Constants.key_pronoun not in participant.vars: # 從上面定義的 treatment pronoun_list 列表中，取出該受試者的 pronoun pid = participant.id_in_session idx_of_participants = Treatment.participant_ids.index(pid) participant.vars[Constants.key_pronoun] = Treatment.pronoun_list[idx_of_participants] pronoun = participant.vars[Constants.key_pronoun] return pronoun available_speech_speed_list = [0.7, 0.8, 0.9] speech_speed_list = [] def prepare_speech_speed_list(all_players): # 事先準備好所有受試者的 speech_speed 們 Treatment.speech_speed_list.clear() Treatment.prepare_participant_ids_if_needed(all_players) for each_speed in Treatment.available_speech_speed_list: count = int(len(Treatment.participant_ids) / len(Treatment.available_speech_speed_list)) # 例如有5人，平均分配兩組 treatment，那就是 5/2 = 2（不取餘數） Treatment.speech_speed_list.extend( [each_speed] * count ) # “python - Create list of single item repeated N times - Stack Overflow” https://stackoverflow.com/questions/3459098/create-list-of-single-item-repeated-n-times while len(Treatment.speech_speed_list) < len(Treatment.participant_ids): # 若因前面 for 迴圈中不能整除，也就是 `speech_speed_list` 不能完平均地分配給所有 participants，則最後會不夠，在此補上。（如果 `speed_available_list` 就兩個，那最多就差1，也就是最多會差到(count - 1)個） Treatment.speech_speed_list.append(random.choice(Treatment.available_speech_speed_list)) # 剩下的就隨機挑選 random.shuffle(Treatment.speech_speed_list) # 打亂原有 list 順序 def get_speech_speed(player): participant = player.participant # lazy loading: 若不存在就取得並存起來，若已存在就直接取出 if Constants.key_speech_speed not in participant.vars: # 從上面定義的 treatment speech_speed_list 列表中，取出該受試者的 speech_speed pid = participant.id_in_session idx_of_participants = Treatment.participant_ids.index(pid) participant.vars[Constants.key_speech_speed] = Treatment.speech_speed_list[idx_of_participants] speech_speed_list = participant.vars[Constants.key_speech_speed] return speech_speed_list class Constants(BaseConstants): name_in_url = 'EconExp1_questionaire' players_per_group = None num_rounds = 24 # 這裡代表最大可能 rounds 數，實際的 rounds 數請參考 actual_num_rounds() key_q_params_pairs = 'questionaire_parameters_pairs' key_selected_q = 'selected_questionaire' key_pronoun = 'treatment_pronoun' key_speech_speed = 'treatment_speech_speed' def actual_num_rounds(): return min(Constants.num_rounds, len(WaitingPeriod.list) * len(GainedAmount.list)) # 取其中最小的 class OptionOfGetMoney(Enum): OPTION_NOW = '選擇今天的報酬' OPTION_FUTURE = '選擇未來的報酬' def formatted_option(player, option_enum): return Treatment.get_pronoun(player) + option_enum.value class Subsession(BaseSubsession): @staticmethod def load_from_session_config_if_needed(session_config): # 從 session config 讀取（預設定義在 settings.py 中 SESSION_CONFIGS，但可在網頁的「Create a new session / Configure session」 中修改） if len(WaitingPeriod.list) == 0: json_string = session_config['available_waiting_periods'] WaitingPeriod.list = json.loads(json_string) if len(GainedAmount.list) == 0: json_string = session_config['available_gained_amounts'] GainedAmount.list = json.loads(json_string) def creating_session(self): Subsession.load_from_session_config_if_needed(self.session.config) Treatment.prepare_pronoun_list(self.get_players()) Treatment.prepare_speech_speed_list(self.get_players()) for p in self.get_players(): p.treatment_pronoun = Treatment.get_pronoun(p) p.treatment_speech_speed = Treatment.get_speech_speed(p) class Group(BaseGroup): pass class Player(BasePlayer): # 幾週後 (hidden) waiting_period = models.IntegerField() # 獲得的報償 (hidden) gained_amount = models.IntegerField() # treatement 主詞（hidden） treatment_pronoun = models.StringField(initial = '') # treatement 語音速度（hidden） treatment_speech_speed = models.FloatField(initial = 0) get_money_now_or_future = models.StringField( label = '請選擇您要今天或未來的報酬', widget = widgets.RadioSelect, ) def get_money_now_or_future_choices(self): return [ ['now', OptionOfGetMoney.formatted_option(self, OptionOfGetMoney.OPTION_NOW)], ['future', OptionOfGetMoney.formatted_option(self, OptionOfGetMoney.OPTION_FUTURE)], ] # 聽了幾次，單位為次數 (hidden，根據使用者行為紀錄) num_listen_times = models.IntegerField(initial = 0) # 決策時長，單位為秒數 (hidden，根據使用者行為紀錄) decision_duration = models.FloatField(initial = 0) # 是否為最後電腦選中的 round (hidden，最後一回合會抽出並寫入) is_selected = models.BooleanField(initial = False)