synth-ai 0.1.0.dev38__py3-none-any.whl → 0.1.0.dev49__py3-none-any.whl

synth_ai-0.1.0.dev38.dist-info/RECORD

@@ -1,67 +0,0 @@
-private_tests/try_synth_sdk.py,sha256=vk4lUEfpQfLACFl6Qw468t_lsuYxuoIIr05WRgWKGKY,24
-public_tests/test_agent.py,sha256=OGmUsd-f0ZRL0xIwJ0MO_dPYQOshNdnmmESxfTSytDM,22552
-public_tests/test_all_structured_outputs.py,sha256=bIcchimaVkq8q8D-GKO25d1_SauTFh4a0c3OEKh3_8o,6524
-public_tests/test_anthropic_structured_outputs.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-public_tests/test_deepseek_structured_outputs.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-public_tests/test_deepseek_tools.py,sha256=MxEaiT_zinuMPeylqfNMsD11zRaMEwY0Fi28bg5op0A,1895
-public_tests/test_gemini_output.py,sha256=704NCnxNepYjUxJj3eEms6zHRCps2PSaR8A-lcsQxb4,6062
-public_tests/test_gemini_structured_outputs.py,sha256=yKa3CDVJxE_Vb2BbVROje83Pb35MBusF0Nb-ttWbqS8,4001
-public_tests/test_models.py,sha256=QGevBfBuQzwyKw1ez34igDyJpMTBVOc3meW6yqFE-bM,5853
-public_tests/test_openai_structured_outputs.py,sha256=oIhdZ2QVLmn0LaqBpCP3Qhbn2KHJv633DGn6u9Ousak,3999
-public_tests/test_reasoning_effort.py,sha256=w4dIiEaEU8gnfAmjrpCC5y-c9w-eH9NzFjwUHe2deyg,3089
-public_tests/test_reasoning_models.py,sha256=Vr4sFRYcrYOBAZMFz2a0fZQqa-WjRwbtwc6lXy6bF4I,2897
-public_tests/test_recursive_structured_outputs.py,sha256=rrqzsU5ExNt-m_wu9j_fkbHiEsAtbKEK66uK5Ub2ojs,6296
-public_tests/test_structured.py,sha256=rftVwvYgMSHkRZM1WUJzga5Uvl9hmc5OpXzBshEXNF0,3740
-public_tests/test_structured_outputs.py,sha256=9SFpH4RQ6nRcphBVmELRNSvhRjYaJBu_z-r6xqKAYpg,4213
-public_tests/test_synth_sdk.py,sha256=jqJHKpvBn9qj21P76z9onXfPg88jyUmBTKmdvCsQMk8,14885
-public_tests/test_text.py,sha256=UyPZ0ci-XBjK35tAeV0kN1X8Njf-0pHfEPZhsWDZ0-c,4072
-public_tests/test_tools.py,sha256=QBwJ70dmPCm27BEwbNaZXXAf8DJxObsfwFX1rlBcYME,10904
-synth_ai/__init__.py,sha256=tX_fcK8u64BoPEboRa3dIKK_WpLy5KAxL2Ucl-l0xVg,147
-synth_ai/zyk/__init__.py,sha256=kGMD-drlBVdsyT-QFODMwaZUtxPCJ9mg58GKQUvFqo0,134
-synth_ai/zyk/lms/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-synth_ai/zyk/lms/config.py,sha256=CcN5NL99j0UZubyGo-MUfbPD3pWosAMju_sqgfvqLVY,201
-synth_ai/zyk/lms/caching/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-synth_ai/zyk/lms/caching/constants.py,sha256=fPi3x9p-yRdvixMSIyclvmwmwCRliXLXQjEm6dRnG8s,52
-synth_ai/zyk/lms/caching/dbs.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-synth_ai/zyk/lms/caching/ephemeral.py,sha256=pNMG5Rzzp2m0Ln1UYmWxz1qbXwq3iNIrhjYAS0yO3ZE,2370
-synth_ai/zyk/lms/caching/handler.py,sha256=4h4Kywf0_-WohE1RxBt4cqPo-kHRjZv-2K50WWO91V4,5050
-synth_ai/zyk/lms/caching/initialize.py,sha256=zZls6RKAax6Z-8oJInGaSg_RPN_fEZ6e_RCX64lMLJw,416
-synth_ai/zyk/lms/caching/persistent.py,sha256=ZaY1A9qhvfNKzcAI9FnwbIrgMKvVeIfb_yCyl3M8dxE,2860
-synth_ai/zyk/lms/core/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-synth_ai/zyk/lms/core/all.py,sha256=wakK0HhvYRuaQZmxClURyNf3vUkTbm3OABw3TgpMjOQ,1185
-synth_ai/zyk/lms/core/exceptions.py,sha256=K0BVdAzxVIchsvYZAaHEH1GAWBZvpxhFi-SPcJOjyPQ,205
-synth_ai/zyk/lms/core/main.py,sha256=NNPd4wwpgscFtCCrVPgz6gcrg7kOTSKsBFhldV0kwv0,12502
-synth_ai/zyk/lms/core/vendor_clients.py,sha256=C4ICuczCG2yRpDbrraT0LUoaPFYytuetfJLqhgvGn8A,2785
-synth_ai/zyk/lms/cost/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-synth_ai/zyk/lms/cost/monitor.py,sha256=cSKIvw6WdPZIRubADWxQoh1MdB40T8-jjgfNUeUHIn0,5
-synth_ai/zyk/lms/cost/statefulness.py,sha256=TOsuXL8IjtKOYJ2aJQF8TwJVqn_wQ7AIwJJmdhMye7U,36
-synth_ai/zyk/lms/structured_outputs/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-synth_ai/zyk/lms/structured_outputs/handler.py,sha256=Y7qQ8VReofLKDX6M7L5OXBUmTyHw6bWEfYz0jqvQIZ0,16969
-synth_ai/zyk/lms/structured_outputs/inject.py,sha256=Fy-zDeleRxOZ8ZRM6IuZ6CP2XZnMe4K2PEn4Q9c_KPY,11777
-synth_ai/zyk/lms/structured_outputs/rehabilitate.py,sha256=GuIhzsb7rTvwgn7f9I9omNnXBz5Me_qrtNYcTWzw5_U,7909
-synth_ai/zyk/lms/tools/base.py,sha256=j7wYb1xAvaAm3qVrINphgUhGS-UjZmRpbouseQYgh7A,3228
-synth_ai/zyk/lms/vendors/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-synth_ai/zyk/lms/vendors/base.py,sha256=aK4PEtkMLt_o3qD22kW-x3HJUEKdIk06zlH4kX0VkAE,760
-synth_ai/zyk/lms/vendors/constants.py,sha256=3CCq45otD80yaLts5sFHvPgLCQNkcjHkc9cqOQ0zH4Y,320
-synth_ai/zyk/lms/vendors/openai_standard.py,sha256=A8fRaazdaccFW24De2dDyiH2kdjhf9SAlQhLmAQuZjA,12934
-synth_ai/zyk/lms/vendors/retries.py,sha256=m-WvAiPix9ovnO2S-m53Td5VZDWBVBFuHuSK9--OVxw,38
-synth_ai/zyk/lms/vendors/core/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-synth_ai/zyk/lms/vendors/core/anthropic_api.py,sha256=cgSPFgR9YM8eO1if92fch6pVv0Dxk3xigSXExGSdu-4,15329
-synth_ai/zyk/lms/vendors/core/gemini_api.py,sha256=gsWudX1RP2a4mjuyWZXT3LUek-UMUYhCfE5TMYxp0nA,11530
-synth_ai/zyk/lms/vendors/core/mistral_api.py,sha256=-EMPBEIoYxxDMxukmcmKL8AGAHPNYe4w-76gsPtmrhk,11860
-synth_ai/zyk/lms/vendors/core/openai_api.py,sha256=GDCHIc0kpCnNPj2oW8RE3Cj2U_HcbXzzA5JV1ArAQlE,6600
-synth_ai/zyk/lms/vendors/local/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-synth_ai/zyk/lms/vendors/local/ollama.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-synth_ai/zyk/lms/vendors/supported/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-synth_ai/zyk/lms/vendors/supported/deepseek.py,sha256=BElW0NGpkSA62wOqzzMtDw8XR36rSNXK5LldeHJkQrc,2430
-synth_ai/zyk/lms/vendors/supported/groq.py,sha256=Fbi7QvhdLx0F-VHO5PY-uIQlPR0bo3C9h1MvIOx8nz0,388
-synth_ai/zyk/lms/vendors/supported/ollama.py,sha256=K30VBFRTd7NYyPmyBVRZS2sm0UB651AHp9i3wd55W64,469
-synth_ai/zyk/lms/vendors/supported/together.py,sha256=Ni_jBqqGPN0PkkY-Ew64s3gNKk51k3FCpLSwlNhKbf0,342
-synth_ai-0.1.0.dev38.dist-info/licenses/LICENSE,sha256=ynhjRQUfqA_RdGRATApfFA_fBAy9cno04sLtLUqxVFM,1069
-tests/test_agent.py,sha256=CjPPWuMWC_TzX1DkDald-bbAxgjXE-HPQvFhq2B--5k,22363
-tests/test_recursive_structured_outputs.py,sha256=Ne-9XwnOxN7eSpGbNHOpegR-sRj589I84T6y8Z_4QnA,5781
-tests/test_structured_outputs.py,sha256=J7sfbGZ7OeB5ONIKpcCTymyayNyAdFfGokC1bcUrSx0,3651
-synth_ai-0.1.0.dev38.dist-info/METADATA,sha256=JHm6Ol6dT8wqnCf-Iq3dQJFxII1oFAMFHDkJY9tDm4o,2702
-synth_ai-0.1.0.dev38.dist-info/WHEEL,sha256=CmyFI0kx5cdEMTLiONQRbGQwjIoR1aIYB7eCAQ4KPJ0,91
-synth_ai-0.1.0.dev38.dist-info/top_level.txt,sha256=5GzJO9j-KbJ_4ppxhmCUa_qdhHM4-9cHHNU76yAI8do,42
-synth_ai-0.1.0.dev38.dist-info/RECORD,,

synth_ai-0.1.0.dev38.dist-info/top_level.txt

@@ -1,4 +0,0 @@
-private_tests
-public_tests
-synth_ai
-tests

tests/test_agent.py

@@ -1,538 +0,0 @@
-import unittest
-from typing import List, Union
-
-from pydantic import BaseModel
-
-from synth_ai.zyk.lms.core.main import LM
-
-
-class UnitTestDict(BaseModel):
-    test_description: str
-    input_names: List[str]
-    input_types: List[str]
-    stringified_input_values: List[str]
-    assertion_condition: str
-    assertion_type: str  # Consider using Literal for specific assertion types
-
-
-class ActionArgument(BaseModel):
-    key: str
-    value: Union[str, int, float, bool, UnitTestDict]
-
-
-class ReAct(BaseModel):
-    reasoning: str
-    action_name: str
-    action_args: List[ActionArgument]  # Dict#[str, Dict]
-
-
-system = """
-<System Message> 
-# Premise
-You are a software engineer
-Here is some information about this setting
-<Setting Information>
-You are working to solve a computer science problem. You will need to submit a solution to the problem, which will be tested against a suite of hidden unit tests.
-</Setting Information>
-<Actions Available>
-<edit_submission>
-<action_context>
-Edit the submission code. Use this when you want to make changes to the current solution.
-</action_context>
-<action_arg_spec>
-{'first_line': <class 'int'>, 'last_line': <class 'int'>, 'new_code': <class 'str'>}
-</action_arg_spec>
-<action_description>
-Edit the submission code
-</action_description>
-
-</edit_submission>
-<add_submission>
-<action_context>
-Add the submission code. Use this when you want to start from scratch with a new solution.
-</action_context>
-<action_arg_spec>
-{'submission': <class 'str'>}
-</action_arg_spec>
-<action_description>
-Add the submission code
-</action_description>
-
-</add_submission>
-<add_unit_test>
-<action_context>
-Add a unit test. The unit test information you submit must be in the format of a BCBUnitTest: 
-
-class BCBUnitTest(BaseModel):
-    test_description: str
-    input_names: List[str]
-    input_types: List[str]
-    input_values: List[Any]
-    assertion_condition: str
-    assertion_type: Literal["assertTrue", "assertRaises"] = "assertTrue"
-
-
- It will be parsed via BCBUnitTest(**unit_test_dict)
-
-
-
-# Some various notes:
-1. If an input should be of a type defined by a specific package, add the package name/alias to the type. E.g. "np.ndarray" or "pd.DataFrame". You still should fully define the value for the input_value field e.g. "pd.DataFrame({'a': [1, 2, 3]})"
-
-2. Unit tests will be compiled from the BCBUnitTest class as follows:
-    A. For AssertTrue type tests, the test will be compiled as follows:
-    ```python
-    def test_case(self):
-        # {{self.test_description}}
-
-        {{defs}}
-        result = {{function_name}}(**{{{{args}}}}})
-        self.{{self.assertion_type}}({{self.assertion_condition}})
-    ```
-    B. For AssertRaises type tests, the test will be compiled as follows:
-
-    ```python
-    def test_case(self):
-        # {{self.test_description}}
-        {{defs}}
-        with self.{{self.assertion_type}}({{self.assertion_condition}}):
-            {{function_name}}(**{{{{args}}}}})
-    ```
-
-    Provide information accordingly.
-
-</action_context>
-<action_arg_spec>
-{'unit_test_name': <class 'str'>, 'unit_test_dict': typing.Dict}
-</action_arg_spec>
-<action_description>
-Add a unit test
-</action_description>
-
-</add_unit_test>
-<remove_unit_test>
-<action_context>
-Remove a unit test
-</action_context>
-<action_arg_spec>
-{'unit_test_name': <class 'str'>}
-</action_arg_spec>
-<action_description>
-Remove a unit test
-</action_description>
-
-</remove_unit_test>
-<test_submission>
-<action_context>
-Test the submission
-</action_context>
-<action_arg_spec>
-{}
-</action_arg_spec>
-<action_description>
-Test the submission
-</action_description>
-
-</test_submission>
-<submit_solution>
-<action_context>
-Submit the solution
-</action_context>
-<action_arg_spec>
-{}
-</action_arg_spec>
-<action_description>
-Submit the solution
-</action_description>
-
-</submit_solution>
-
-</Actions Available>
-You'll be given your past actions/thoughts, along with recent raw observations from the environment
-The environment one step in the past is your current environment.
-
-# Objective
-Please complete the problem by drafting a solution, creating unit tests, improving the solution, and submitting the solution.
-
-# Constraints
-You will be given a code_prompt_for_answer, which contains imports and the function signature. Your solution must comprise code that can be appended to code_prompt_for_answer and run as a single script.
- 
-"""
-
-user = """
-<User Message> 
-# Recent Actions / Thoughts
-
-# Recent Observations
-<1 environment step(s) in the past>{'action_result': None, 'environment_state': {'question': 'import pandas as pd\nimport numpy as np\n\n# Constants\nCOLUMNS = [\'column1\', \'column2\', \'column3\', \'column4\', \'column5\']\n\ndef task_func(df, dct):\n    '''\n    Replace certain values in a DataFrame with a dictionary mapping and calculate the Pearson correlation coefficient between each pair of columns.\n\n    Parameters:\n    df (DataFrame): The input DataFrame, containing numeric or categorical data.\n    dct (dict): A dictionary for replacing values in df, where keys are existing values and values are new values.\n\n    Returns:\n    DataFrame: A DataFrame with the correlation coefficients between each pair of columns. The format of the DataFrame is a square matrix with column and index labels matching the columns of the input DataFrame.\n    \n    Requirements:\n    - pandas\n    - numpy\n    \n    Note:\n    - This function operates on DataFrames containing numeric or categorical data that can be replaced with numeric values, as correlation calculations require numeric data.\n    - This function using pearson method to calculate the correlation matrix.\n    \n    Raises:\n    - This function will raise a ValueError is input df is not a DataFrame.\n        \n    Example:\n    >>> df = pd.DataFrame({\'A\': [1, 2, 3], \'B\': [4, 5, 6]})\n    >>> dct = {1: 10, 2: 20, 3: 30, 4: 40, 5: 50, 6: 60}\n    >>> correlation_matrix = task_func(df, dct)\n    >>> correlation_matrix.shape == (2, 2)\n    True\n    >>> np.allclose(correlation_matrix, np.array([[1.0, 1.0], [1.0, 1.0]]))\n    True\n    '''\n', 'code_prompt_for_answer': "import pandas as pd\nimport numpy as np\n# Constants\nCOLUMNS = ['column1', 'column2', 'column3', 'column4', 'column5']\ndef task_func(df, dct):\n", 'unit_tests_you_have_written': {}, 'current_solution': ''}}</1 environment step(s) in the past>
-
-Your next actions / thought:
-"""
-
-
-###
-
-hard_system = """
-# Premise
-You are a software engineer
-Here is some information about this setting
-<Setting Information>
-You are working to solve a computer science problem. You will need to submit a solution to the problem, which will be tested against a suite of hidden unit tests.
-</Setting Information>
-<Actions Available>
-<edit_submission>
-<action_context>
-Edit the submission code. Use this when you want to make changes to the current solution.
-</action_context>
-<action_arg_spec>
-{'first_line': <class 'int'>, 'last_line': <class 'int'>, 'new_code': <class 'str'>}
-</action_arg_spec>
-<action_description>
-Edit the submission code
-</action_description>
-
-</edit_submission>
-<add_submission>
-<action_context>
-Add the submission code. Use this when you want to start from scratch with a new solution.
-</action_context>
-<action_arg_spec>
-{'submission': <class 'str'>}
-</action_arg_spec>
-<action_description>
-Add the submission code
-</action_description>
-
-</add_submission>
-<add_unit_test>
-<action_context>
-Add a unit test. The unit test information you submit must be in the format of a BCBUnitTest: 
-
-class BCBUnitTest(BaseModel):
-    test_description: str
-    input_names: List[str]
-    input_types: List[str]
-    input_values: List[Any]
-    assertion_condition: str
-    assertion_type: Literal["assertTrue", "assertRaises"] = "assertTrue"
-
-
- It will be parsed via BCBUnitTest(**unit_test_dict)
-
-
-
-# Some various notes:
-1. If an input should be of a type defined by a specific package, add the package name/alias to the type. E.g. "np.ndarray" or "pd.DataFrame". You still should fully define the value for the input_value field e.g. "pd.DataFrame({'a': [1, 2, 3]})"
-
-2. Unit tests will be compiled from the BCBUnitTest class as follows:
-    A. For AssertTrue type tests, the test will be compiled as follows:
-    ```python
-    def test_case(self):
-        # {{self.test_description}}
-
-        {{defs}}
-        result = {{function_name}}(**{{{{args}}}}})
-        self.{{self.assertion_type}}({{self.assertion_condition}})
-    ```
-    B. For AssertRaises type tests, the test will be compiled as follows:
-
-    ```python
-    def test_case(self):
-        # {{self.test_description}}
-        {{defs}}
-        with self.{{self.assertion_type}}({{self.assertion_condition}}):
-            {{function_name}}(**{{{{args}}}}})
-    ```
-
-    Provide information accordingly.
-
-</action_context>
-<action_arg_spec>
-{'unit_test_name': <class 'str'>, 'unit_test_dict': typing.Dict}
-</action_arg_spec>
-<action_description>
-Add a unit test
-</action_description>
-
-</add_unit_test>
-<remove_unit_test>
-<action_context>
-Remove a unit test
-</action_context>
-<action_arg_spec>
-{'unit_test_name': <class 'str'>}
-</action_arg_spec>
-<action_description>
-Remove a unit test
-</action_description>
-
-</remove_unit_test>
-<test_submission>
-<action_context>
-Test the submission
-</action_context>
-<action_arg_spec>
-{}
-</action_arg_spec>
-<action_description>
-Test the submission
-</action_description>
-
-</test_submission>
-<submit_solution>
-<action_context>
-Submit the solution
-</action_context>
-<action_arg_spec>
-{}
-</action_arg_spec>
-<action_description>
-Submit the solution
-</action_description>
-
-</submit_solution>
-
-</Actions Available>
-You'll be given your past actions/thoughts, along with recent raw observations from the environment
-The environment one step in the past is your current environment.
-
-# Objective
-Please complete the problem by drafting a solution, creating unit tests, improving the solution, and submitting the solution.
-
-# Constraints
-You will be given a code_prompt_for_answer, which contains imports and the function signature. Your solution must comprise code that can be appended to code_prompt_for_answer and run as a single script.
- 
-
-<User Message> 
-# Recent Actions / Thoughts
-
-# Recent Observations
-<1 environment step(s) in the past>{'action_result': None, 'environment_state': {'question': 'import pandas as pd\nimport numpy as np\n\n# Constants\nCOLUMNS = [\'column1\', \'column2\', \'column3\', \'column4\', \'column5\']\n\ndef task_func(df, dct):\n   '''\n    Replace certain values in a DataFrame with a dictionary mapping and calculate the Pearson correlation coefficient between each pair of columns.\n\n    Parameters:\n    df (DataFrame): The input DataFrame, containing numeric or categorical data.\n    dct (dict): A dictionary for replacing values in df, where keys are existing values and values are new values.\n\n    Returns:\n    DataFrame: A DataFrame with the correlation coefficients between each pair of columns. The format of the DataFrame is a square matrix with column and index labels matching the columns of the input DataFrame.\n    \n    Requirements:\n    - pandas\n    - numpy\n    \n    Note:\n    - This function operates on DataFrames containing numeric or categorical data that can be replaced with numeric values, as correlation calculations require numeric data.\n    - This function using pearson method to calculate the correlation matrix.\n    \n    Raises:\n    - This function will raise a ValueError is input df is not a DataFrame.\n        \n    Example:\n    >>> df = pd.DataFrame({\'A\': [1, 2, 3], \'B\': [4, 5, 6]})\n    >>> dct = {1: 10, 2: 20, 3: 30, 4: 40, 5: 50, 6: 60}\n    >>> correlation_matrix = task_func(df, dct)\n    >>> correlation_matrix.shape == (2, 2)\n    True\n    >>> np.allclose(correlation_matrix, np.array([[1.0, 1.0], [1.0, 1.0]]))\n    True\n    '''\n', 'code_prompt_for_answer': "import pandas as pd\nimport numpy as np\n# Constants\nCOLUMNS = ['column1', 'column2', 'column3', 'column4', 'column5']\ndef task_func(df, dct):\n", 'unit_tests_you_have_written': {}, 'current_solution': ''}}</1 environment step(s) in the past>
-
-Your next actions / thought:  
-
-Structured output: reasoning="I need to implement the function 'task_func' that replaces values in a DataFrame based on a dictionary and calculates the Pearson correlation coefficient between the columns. I will also ensure to handle the case where the input is not a DataFrame by raising a ValueError." action_name='edit_submission' action_args=[ActionArgument(key='first_line', value=4), ActionArgument(key='last_line', value=4), ActionArgument(key='new_code', value="    if not isinstance(df, pd.DataFrame):\n        raise ValueError('Input must be a DataFrame')\n    df.replace(dct, inplace=True)\n    return df.corr(method='pearson')")]
-<System Message> 
-# Premise
-You are a software engineer
-Here is some information about this setting
-<Setting Information>
-You are working to solve a computer science problem. You will need to submit a solution to the problem, which will be tested against a suite of hidden unit tests.
-</Setting Information>
-<Actions Available>
-<edit_submission>
-<action_context>
-Edit the submission code. Use this when you want to make changes to the current solution.
-</action_context>
-<action_arg_spec>
-{'first_line': <class 'int'>, 'last_line': <class 'int'>, 'new_code': <class 'str'>}
-</action_arg_spec>
-<action_description>
-Edit the submission code
-</action_description>
-
-</edit_submission>
-<add_submission>
-<action_context>
-Add the submission code. Use this when you want to start from scratch with a new solution.
-</action_context>
-<action_arg_spec>
-{'submission': <class 'str'>}
-</action_arg_spec>
-<action_description>
-Add the submission code
-</action_description>
-
-</add_submission>
-<add_unit_test>
-<action_context>
-Add a unit test. The unit test information you submit must be in the format of a BCBUnitTest: 
-
-class BCBUnitTest(BaseModel):
-    test_description: str
-    input_names: List[str]
-    input_types: List[str]
-    input_values: List[Any]
-    assertion_condition: str
-    assertion_type: Literal["assertTrue", "assertRaises"] = "assertTrue"
-
-
- It will be parsed via BCBUnitTest(**unit_test_dict)
-
-
-
-# Some various notes:
-1. If an input should be of a type defined by a specific package, add the package name/alias to the type. E.g. "np.ndarray" or "pd.DataFrame". You still should fully define the value for the input_value field e.g. "pd.DataFrame({'a': [1, 2, 3]})"
-
-2. Unit tests will be compiled from the BCBUnitTest class as follows:
-    A. For AssertTrue type tests, the test will be compiled as follows:
-    ```python
-    def test_case(self):
-        # {{self.test_description}}
-
-        {{defs}}
-        result = {{function_name}}(**{{{{args}}}}})
-        self.{{self.assertion_type}}({{self.assertion_condition}})
-    ```
-    B. For AssertRaises type tests, the test will be compiled as follows:
-
-    ```python
-    def test_case(self):
-        # {{self.test_description}}
-        {{defs}}
-        with self.{{self.assertion_type}}({{self.assertion_condition}}):
-            {{function_name}}(**{{{{args}}}}})
-    ```
-
-    Provide information accordingly.
-
-</action_context>
-<action_arg_spec>
-{'unit_test_name': <class 'str'>, 'unit_test_dict': typing.Dict}
-</action_arg_spec>
-<action_description>
-Add a unit test
-</action_description>
-
-</add_unit_test>
-<remove_unit_test>
-<action_context>
-Remove a unit test
-</action_context>
-<action_arg_spec>
-{'unit_test_name': <class 'str'>}
-</action_arg_spec>
-<action_description>
-Remove a unit test
-</action_description>
-
-</remove_unit_test>
-<test_submission>
-<action_context>
-Test the submission
-</action_context>
-<action_arg_spec>
-{}
-</action_arg_spec>
-<action_description>
-Test the submission
-</action_description>
-
-</test_submission>
-<submit_solution>
-<action_context>
-Submit the solution
-</action_context>
-<action_arg_spec>
-{}
-</action_arg_spec>
-<action_description>
-Submit the solution
-</action_description>
-
-</submit_solution>
-
-</Actions Available>
-You'll be given your past actions/thoughts, along with recent raw observations from the environment
-The environment one step in the past is your current environment.
-
-# Objective
-Please complete the problem by drafting a solution, creating unit tests, improving the solution, and submitting the solution.
-
-# Constraints
-You will be given a code_prompt_for_answer, which contains imports and the function signature. Your solution must comprise code that can be appended to code_prompt_for_answer and run as a single script.
-"""
-
-hard_user = """
-# Recent Actions / Thoughts
-<1 reasoning step(s) in the past>reasoning="I need to implement the function 'task_func' that replaces values in a DataFrame based on a dictionary and calculates the Pearson correlation coefficient between the columns. I will also ensure to handle the case where the input is not a DataFrame by raising a ValueError." action_name='edit_submission' action_args=[ActionArgument(key='first_line', value=4), ActionArgument(key='last_line', value=4), ActionArgument(key='new_code', value="    if not isinstance(df, pd.DataFrame):\n        raise ValueError('Input must be a DataFrame')\n    df.replace(dct, inplace=True)\n    return df.corr(method='pearson')")]</1 reasoning step(s) in the past>
-# Recent Observations
-<1 environment step(s) in the past>success=True result='Edited submission successfully'</1 environment step(s) in the past>
-<2 environment step(s) in the past>{'action_result': None, 'environment_state': {'question': 'import pandas as pd\nimport numpy as np\n\n# Constants\nCOLUMNS = [\'column1\', \'column2\', \'column3\', \'column4\', \'column5\']\n\ndef task_func(df, dct):\n    '''\n    Replace certain values in a DataFrame with a dictionary mapping and calculate the Pearson correlation coefficient between each pair of columns.\n\n    Parameters:\n    df (DataFrame): The input DataFrame, containing numeric or categorical data.\n    dct (dict): A dictionary for replacing values in df, where keys are existing values and values are new values.\n\n    Returns:\n    DataFrame: A DataFrame with the correlation coefficients between each pair of columns. The format of the DataFrame is a square matrix with column and index labels matching the columns of the input DataFrame.\n    \n    Requirements:\n    - pandas\n    - numpy\n    \n    Note:\n    - This function operates on DataFrames containing numeric or categorical data that can be replaced with numeric values, as correlation calculations require numeric data.\n    - This function using pearson method to calculate the correlation matrix.\n    \n    Raises:\n    - This function will raise a ValueError is input df is not a DataFrame.\n        \n    Example:\n    >>> df = pd.DataFrame({\'A\': [1, 2, 3], \'B\': [4, 5, 6]})\n    >>> dct = {1: 10, 2: 20, 3: 30, 4: 40, 5: 50, 6: 60}\n    >>> correlation_matrix = task_func(df, dct)\n    >>> correlation_matrix.shape == (2, 2)\n    True\n    >>> np.allclose(correlation_matrix, np.array([[1.0, 1.0], [1.0, 1.0]]))\n    True\n    '''\n', 'code_prompt_for_answer': "import pandas as pd\nimport numpy as np\n# Constants\nCOLUMNS = ['column1', 'column2', 'column3', 'column4', 'column5']\ndef task_func(df, dct):\n", 'unit_tests_you_have_written': {}, 'current_solution': ''}}</2 environment step(s) in the past>
-
-Your next actions / thought:
-"""
-
-
-class TestLMStructuredOutputs(unittest.TestCase):
-    # ... existing code ...
-
-    @classmethod
-    def setUpClass(cls):
-        # Initialize LMs for both forced_json and stringified_json modes
-        cls.lm_forced_json = LM(
-            model_name="gpt-4o-mini",
-            formatting_model_name="gpt-4o-mini",
-            temperature=0.0,
-            max_retries="Few",
-            structured_output_mode="forced_json",
-        )
-        cls.lm_stringified_json = LM(
-            model_name="gpt-4o-mini",
-            formatting_model_name="gpt-4o-mini",
-            temperature=0.0,
-            max_retries="Few",
-            structured_output_mode="stringified_json",
-        )
-
-    def test_sync_react_response_content(self):
-        system_message = system
-
-        user_message = user
-
-        for lm in [self.lm_forced_json, self.lm_stringified_json]:
-            with self.subTest(
-                mode=lm.structured_output_handler.handler.structured_output_mode
-            ):
-                result = lm.respond_sync(
-                    system_message=system_message,
-                    user_message=user_message,
-                    response_model=ReAct,
-                )
-                self.assertIsInstance(result, ReAct)
-                self.assertIsInstance(result.reasoning, str)
-                self.assertIsInstance(result.action_name, str)
-                self.assertIsInstance(result.action_args, list)
-                for arg in result.action_args:
-                    self.assertIsInstance(arg, ActionArgument)
-                    self.assertIsInstance(arg.key, str)
-                    # self.assertIsInstance(arg.value, str)
-
-    def test_sync_react_response_hard_content(self):
-        system_message = hard_system
-
-        user_message = hard_user
-
-        for lm in [self.lm_forced_json, self.lm_stringified_json]:
-            with self.subTest(
-                mode=lm.structured_output_handler.handler.structured_output_mode
-            ):
-                result = lm.respond_sync(
-                    system_message=system_message,
-                    user_message=user_message,
-                    response_model=ReAct,
-                )
-                self.assertIsInstance(result, ReAct)
-                self.assertIsInstance(result.reasoning, str)
-                self.assertIsInstance(result.action_name, str)
-                self.assertIsInstance(result.action_args, list)
-                for arg in result.action_args:
-                    self.assertIsInstance(arg, ActionArgument)
-                    self.assertIsInstance(arg.key, str)
-                    # self.assertIsInstance(arg.value, str)
-
-
-# use non-trivial fallback?
-
-if __name__ == "__main__":
-    # Create an instance of the test class
-    test_instance = TestLMStructuredOutputs()
-
-    # Set up the class (this would normally be done by unittest)
-    test_instance.setUpClass()
-
-    # Run the test methods
-    test_instance.test_sync_react_response_content()
-    test_instance.test_sync_react_response_hard_content()
-
-    print("All tests completed.")