edsl 0.1.34__tar.gz → 0.1.34.dev1__tar.gz

{edsl-0.1.34 → edsl-0.1.34.dev1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: edsl
-Version: 0.1.34
+Version: 0.1.34.dev1
 Summary: Create and analyze LLM-based surveys
 Home-page: https://www.expectedparrot.com/
 License: MIT

edsl-0.1.34.dev1/edsl/__version__.py

@@ -0,0 +1 @@
+__version__ = "0.1.34.dev1"

{edsl-0.1.34 → edsl-0.1.34.dev1}/edsl/agents/PromptConstructor.py

@@ -1,7 +1,7 @@
 from __future__ import annotations
 from typing import Dict, Any, Optional, Set
 from collections import UserList
-import pdb
+import enum
 
 from jinja2 import Environment, meta
 
@@ -172,6 +172,7 @@ class PromptConstructor:
         """
         # The user might have passed a custom prompt, which would be stored in _question_instructions_prompt
         if not hasattr(self, "_question_instructions_prompt"):
+
             # Gets the instructions for the question - this is how the question should be answered
             question_prompt = self.question.get_instructions(model=self.model.model)
 
@@ -197,15 +198,14 @@ class PromptConstructor:
                     self.question.question_options = question_options
 
                 # might be getting it from the prior answers
-                if self.prior_answers_dict().get(question_option_key) is not None:
-                    if isinstance(
-                        question_options := self.prior_answers_dict()
-                        .get(question_option_key)
-                        .answer,
-                        list,
-                    ):
-                        question_data["question_options"] = question_options
-                        self.question.question_options = question_options
+                if isinstance(
+                    question_options := self.prior_answers_dict()
+                    .get(question_option_key)
+                    .answer,
+                    list,
+                ):
+                    question_data["question_options"] = question_options
+                    self.question.question_options = question_options
 
             replacement_dict = (
                 {key: f"<see file {key}>" for key in self.scenario_file_keys}

{edsl-0.1.34 → edsl-0.1.34.dev1}/edsl/jobs/Jobs.py

@@ -145,21 +145,14 @@ class Jobs(Base):
         >>> Jobs.example().prompts()
         Dataset(...)
         """
-        from edsl import Coop
-
-        c = Coop()
-        price_lookup = c.fetch_prices()
 
         interviews = self.interviews()
         # data = []
         interview_indices = []
-        question_names = []
+        question_indices = []
         user_prompts = []
         system_prompts = []
         scenario_indices = []
-        agent_indices = []
-        models = []
-        costs = []
         from edsl.results.Dataset import Dataset
 
         for interview_index, interview in enumerate(interviews):
@@ -167,97 +160,23 @@ class Jobs(Base):
                 interview._get_invigilator(question)
                 for question in self.survey.questions
             ]
+            # list(interview._build_invigilators(debug=False))
             for _, invigilator in enumerate(invigilators):
                 prompts = invigilator.get_prompts()
-                user_prompt = prompts["user_prompt"]
-                system_prompt = prompts["system_prompt"]
-                user_prompts.append(user_prompt)
-                system_prompts.append(system_prompt)
-                agent_index = self.agents.index(invigilator.agent)
-                agent_indices.append(agent_index)
+                user_prompts.append(prompts["user_prompt"])
+                system_prompts.append(prompts["system_prompt"])
                 interview_indices.append(interview_index)
-                scenario_index = self.scenarios.index(invigilator.scenario)
-                scenario_indices.append(scenario_index)
-                models.append(invigilator.model.model)
-                question_names.append(invigilator.question.question_name)
-                # cost calculation
-                key = (invigilator.model._inference_service_, invigilator.model.model)
-                relevant_prices = price_lookup[key]
-                inverse_output_price = relevant_prices["output"]["one_usd_buys"]
-                inverse_input_price = relevant_prices["input"]["one_usd_buys"]
-                input_tokens = len(str(user_prompt) + str(system_prompt)) // 4
-                output_tokens = len(str(user_prompt) + str(system_prompt)) // 4
-                cost = input_tokens / float(
-                    inverse_input_price
-                ) + output_tokens / float(inverse_output_price)
-                costs.append(cost)
-
-        d = Dataset(
+                scenario_indices.append(invigilator.scenario)
+                question_indices.append(invigilator.question.question_name)
+        return Dataset(
             [
-                {"user_prompt": user_prompts},
-                {"system_prompt": system_prompts},
                 {"interview_index": interview_indices},
-                {"question_name": question_names},
+                {"question_index": question_indices},
+                {"user_prompt": user_prompts},
                 {"scenario_index": scenario_indices},
-                {"agent_index": agent_indices},
-                {"model": models},
-                {"estimated_cost": costs},
+                {"system_prompt": system_prompts},
             ]
         )
-        return d
-        # if table:
-        #     d.to_scenario_list().print(format="rich")
-        # else:
-        #     return d
-
-    def show_prompts(self) -> None:
-        """Print the prompts."""
-        self.prompts().to_scenario_list().print(format="rich")
-
-    def estimate_job_cost(self):
-        from edsl import Coop
-
-        c = Coop()
-        price_lookup = c.fetch_prices()
-
-        prompts = self.prompts()
-
-        text_len = 0
-        for prompt in prompts:
-            text_len += len(str(prompt))
-
-        input_token_aproximations = text_len // 4
-
-        aproximation_cost = {}
-        total_cost = 0
-        for model in self.models:
-            key = (model._inference_service_, model.model)
-            relevant_prices = price_lookup[key]
-            inverse_output_price = relevant_prices["output"]["one_usd_buys"]
-            inverse_input_price = relevant_prices["input"]["one_usd_buys"]
-
-            aproximation_cost[key] = {
-                "input": input_token_aproximations / float(inverse_input_price),
-                "output": input_token_aproximations / float(inverse_output_price),
-            }
-            ##TODO curenlty we approximate the number of output tokens with the number
-            # of input tokens. A better solution will be to compute the quesiton answer options length and sum them
-            # to compute the output tokens
-
-            total_cost += input_token_aproximations / float(inverse_input_price)
-            total_cost += input_token_aproximations / float(inverse_output_price)
-
-        # multiply_factor = len(self.agents or [1]) * len(self.scenarios or [1])
-        multiply_factor = 1
-        out = {
-            "input_token_aproximations": input_token_aproximations,
-            "models_costs": aproximation_cost,
-            "estimated_total_cost": total_cost * multiply_factor,
-            "multiply_factor": multiply_factor,
-            "single_config_cost": total_cost,
-        }
-
-        return out
 
     @staticmethod
     def _get_container_class(object):

{edsl-0.1.34 → edsl-0.1.34.dev1}/edsl/jobs/interviews/Interview.py

@@ -3,7 +3,6 @@
 from __future__ import annotations
 import asyncio
 from typing import Any, Type, List, Generator, Optional, Union
-import copy
 
 from tenacity import (
     retry,
@@ -100,17 +99,15 @@ class Interview(InterviewStatusMixin):
 
         """
         self.agent = agent
-        # what I would like to do
-        self.survey = copy.deepcopy(survey)  # survey  copy.deepcopy(survey)
-        # self.survey = survey
+        self.survey = survey
         self.scenario = scenario
         self.model = model
         self.debug = debug
         self.iteration = iteration
         self.cache = cache
-        self.answers: dict[
-            str, str
-        ] = Answers()  # will get filled in as interview progresses
+        self.answers: dict[str, str] = (
+            Answers()
+        )  # will get filled in as interview progresses
         self.sidecar_model = sidecar_model
 
         # self.stop_on_exception = False
@@ -431,11 +428,11 @@ class Interview(InterviewStatusMixin):
         """
         current_question_index: int = self.to_index[current_question.question_name]
 
-        next_question: Union[
-            int, EndOfSurvey
-        ] = self.survey.rule_collection.next_question(
-            q_now=current_question_index,
-            answers=self.answers | self.scenario | self.agent["traits"],
+        next_question: Union[int, EndOfSurvey] = (
+            self.survey.rule_collection.next_question(
+                q_now=current_question_index,
+                answers=self.answers | self.scenario | self.agent["traits"],
+            )
         )
 
         next_question_index = next_question.next_q

{edsl-0.1.34 → edsl-0.1.34.dev1}/edsl/jobs/runners/JobsRunnerAsyncio.py

@@ -173,19 +173,19 @@ class JobsRunnerAsyncio:
 
         prompt_dictionary = {}
         for answer_key_name in answer_key_names:
-            prompt_dictionary[
-                answer_key_name + "_user_prompt"
-            ] = question_name_to_prompts[answer_key_name]["user_prompt"]
-            prompt_dictionary[
-                answer_key_name + "_system_prompt"
-            ] = question_name_to_prompts[answer_key_name]["system_prompt"]
+            prompt_dictionary[answer_key_name + "_user_prompt"] = (
+                question_name_to_prompts[answer_key_name]["user_prompt"]
+            )
+            prompt_dictionary[answer_key_name + "_system_prompt"] = (
+                question_name_to_prompts[answer_key_name]["system_prompt"]
+            )
 
         raw_model_results_dictionary = {}
         for result in valid_results:
             question_name = result.question_name
-            raw_model_results_dictionary[
-                question_name + "_raw_model_response"
-            ] = result.raw_model_response
+            raw_model_results_dictionary[question_name + "_raw_model_response"] = (
+                result.raw_model_response
+            )
             raw_model_results_dictionary[question_name + "_cost"] = result.cost
             one_use_buys = (
                 "NA"

{edsl-0.1.34 → edsl-0.1.34.dev1}/edsl/language_models/LanguageModel.py

@@ -490,7 +490,7 @@ class LanguageModel(
                 "user_prompt": user_prompt,
                 "system_prompt": system_prompt,
                 "files_list": files_list
-                # **({"encoded_image": encoded_image} if encoded_image else {}),
+                #**({"encoded_image": encoded_image} if encoded_image else {}),
             }
             # response = await f(**params)
             response = await asyncio.wait_for(f(**params), timeout=TIMEOUT)
@@ -534,7 +534,7 @@ class LanguageModel(
         system_prompt: str,
         cache: "Cache",
         iteration: int = 1,
-        files_list: Optional[List["File"]] = None,
+        files_list: Optional[List['File']] = None,
     ) -> dict:
         """Get response, parse, and return as string.
 

{edsl-0.1.34 → edsl-0.1.34.dev1}/edsl/language_models/utilities.py

@@ -40,10 +40,8 @@ def create_language_model(
         _tpm = 1000000000000
 
         async def async_execute_model_call(
-            self,
-            user_prompt: str,
-            system_prompt: str,
-            files_list: Optional[List[Any]] = None,
+            self, user_prompt: str, system_prompt: str, 
+            files_list: Optional[List[Any]] = None
         ) -> dict[str, Any]:
             question_number = int(
                 user_prompt.split("XX")[1]

{edsl-0.1.34 → edsl-0.1.34.dev1}/edsl/questions/QuestionBase.py

@@ -82,8 +82,7 @@ class QuestionBase(
         if not hasattr(self, "_fake_data_factory"):
             from polyfactory.factories.pydantic_factory import ModelFactory
 
-            class FakeData(ModelFactory[self.response_model]):
-                ...
+            class FakeData(ModelFactory[self.response_model]): ...
 
             self._fake_data_factory = FakeData
         return self._fake_data_factory

{edsl-0.1.34 → edsl-0.1.34.dev1}/edsl/questions/QuestionMultipleChoice.py

@@ -120,9 +120,9 @@ class QuestionMultipleChoice(QuestionBase):
 
     question_type = "multiple_choice"
     purpose = "When options are known and limited"
-    question_options: Union[
-        list[str], list[list], list[float], list[int]
-    ] = QuestionOptionsDescriptor()
+    question_options: Union[list[str], list[list], list[float], list[int]] = (
+        QuestionOptionsDescriptor()
+    )
     _response_model = None
     response_validator_class = MultipleChoiceResponseValidator
 

{edsl-0.1.34 → edsl-0.1.34.dev1}/edsl/questions/templates/numerical/answering_instructions.jinja

@@ -1,6 +1,7 @@
 This question requires a numerical response in the form of an integer or decimal (e.g., -12, 0, 1, 2, 3.45, ...).
 Respond with just your number on a single line.
 If your response is equivalent to zero, report '0'
+If you cannot determine the answer, report 'None'
 
 {% if include_comment %}
 After the answer, put a comment explaining your choice on the next line.

{edsl-0.1.34 → edsl-0.1.34.dev1}/edsl/scenarios/FileStore.py

@@ -263,6 +263,7 @@ class FileStore(Scenario):
 
 
 class CSVFileStore(FileStore):
+
     @classmethod
     def example(cls):
         from edsl.results.Results import Results
@@ -281,6 +282,7 @@ class CSVFileStore(FileStore):
 
 
 class PDFFileStore(FileStore):
+
     def view(self):
         pdf_path = self.to_tempfile()
         print(f"PDF path: {pdf_path}")  # Print the path to ensure it exists
@@ -356,6 +358,7 @@ class PDFFileStore(FileStore):
 
 
 class PNGFileStore(FileStore):
+
     @classmethod
     def example(cls):
         import textwrap
@@ -379,6 +382,7 @@ class PNGFileStore(FileStore):
 
 
 class SQLiteFileStore(FileStore):
+
     @classmethod
     def example(cls):
         import sqlite3
@@ -401,6 +405,7 @@ class SQLiteFileStore(FileStore):
 
 
 class HTMLFileStore(FileStore):
+
     @classmethod
     def example(cls):
         import tempfile

{edsl-0.1.34 → edsl-0.1.34.dev1}/edsl/surveys/base.py

@@ -36,10 +36,6 @@ class EndOfSurveyParent:
         """
         return self
 
-    def __deepcopy__(self, memo):
-        # Return the same instance when deepcopy is called
-        return self
-
     def __radd__(self, other):
         """Add the object to another object.
 

{edsl-0.1.34 → edsl-0.1.34.dev1}/pyproject.toml

@@ -1,19 +1,28 @@
 [build-system]
 build-backend = "poetry.core.masonry.api"
-requires = ["poetry-core", "setuptools<72.0"]
+requires = [ "poetry-core", "setuptools<72.0",]
 
 [tool.poetry]
-authors = ["Apostolos Filippas <apostolos@expectedparrot.com>", "John Horton <john@expectedparrot.com>", "Robin Horton <robin@expectedparrot.com>"]
-classifiers = ["Topic :: Scientific/Engineering :: Artificial Intelligence", "Topic :: Software Development :: Libraries :: Application Frameworks", "Topic :: Software Development :: Libraries :: Python Modules"]
+authors = [ "Apostolos Filippas <apostolos@expectedparrot.com>", "John Horton <john@expectedparrot.com>", "Robin Horton <robin@expectedparrot.com>",]
+classifiers = [ "Topic :: Scientific/Engineering :: Artificial Intelligence", "Topic :: Software Development :: Libraries :: Application Frameworks", "Topic :: Software Development :: Libraries :: Python Modules",]
 description = "Create and analyze LLM-based surveys"
 documentation = "https://docs.expectedparrot.com"
 homepage = "https://www.expectedparrot.com/"
-include = ["edsl/questions/templates/**/*"]
-keywords = ["LLM", "social science", "surveys", "user research"]
+include = [ "edsl/questions/templates/**/*",]
+keywords = [ "LLM", "social science", "surveys", "user research",]
 license = "MIT"
 name = "edsl"
 readme = "README.md"
-version = "0.1.34"
+version = "0.1.34.dev1"
+[[tool.poetry.packages]]
+include = "edsl"
+
+[tool.tomlsort]
+all = true
+in_place = true
+spaces_before_inline_comment = 2
+spaces_indent_inline_array = 4
+trailing_comma_inline_array = true
 
 [tool.poetry.dependencies]
 python = ">=3.9.1,<3.13"
@@ -49,9 +58,12 @@ urllib3 = ">=1.25.4,<1.27"
 google-generativeai = "^0.8.2"
 
 [tool.poetry.dependencies.black]
-extras = ["jupyter"]
+extras = [ "jupyter",]
 version = "^24.4.2"
 
+[tool.tomlsort.overrides."tool.poetry.dependencies"]
+table_keys = false
+
 [tool.poetry.group.dev.dependencies]
 coverage = "^7.3.3"
 fastapi = "^0.112.1"
@@ -77,16 +89,3 @@ sphinx-rtd-theme = "^2.0.0"
 toml = "^0.10.2"
 toml-sort = "^0.23.1"
 uvicorn = "^0.30.6"
-
-[[tool.poetry.packages]]
-include = "edsl"
-
-[tool.tomlsort]
-all = true
-in_place = true
-spaces_before_inline_comment = 2
-spaces_indent_inline_array = 4
-trailing_comma_inline_array = true
-
-[tool.tomlsort.overrides."tool.poetry.dependencies"]
-table_keys = false

edsl-0.1.34/edsl/__version__.py

@@ -1 +0,0 @@
-__version__ = "0.1.34"