consistent import of cleandoc and addressed comments

Signed-off-by: Amit Sharma <[email protected]>

Author: amit-sharma

pywhyllm/suggesters/identification_suggester.py

@@ -4,7 +4,7 @@
 import guidance
 from guidance import system, user, assistant, gen
 import re
-import inspect
+from inspect import cleandoc
 
 class IdentificationSuggester(IdentifierProtocol):
     CONTEXT: str = """causal mechanisms"""
@@ -204,7 +204,7 @@ def request_mediators(
                 influencing athe {outcome}. Where a causal model is a conceptual model that describes the causal mechanisms 
                 of a system. You will do this by by answering questions about cause and effect and using your domain knowledge 
                 in {domain_expertise}. Follow the next two steps, and complete the first one before moving on to the second:"""
-                    lm += inspect.cleandoc(prompt_str)
+                    lm += cleandoc(prompt_str)
 
                 with user():
                     prompt_str = f"""(1) From your perspective as an expert in {domain_expertise}, think step by step as you consider the factors 
@@ -224,7 +224,7 @@ def request_mediators(
                   the tags <mediating_factor>factor_name</mediating_factor>. Where factor_name is one of the items within the 
                   factor_names list. If a factor does not have a high likelihood of mediating, then do not wrap the factor with 
                   any tags. Your step by step answer as an in {domain_expertise}:"""
-                    lm += inspect.cleandoc(prompt_str)
+                    lm += cleandoc(prompt_str)
 
                 with assistant():
                     lm += gen("output")
@@ -322,7 +322,7 @@ def request_ivs(
                         causal mechanisms of a system. You will do this by by answering questions about cause and effect and using 
                         your domain knowledge in {domain_expertise}. Follow the next two steps, and complete the first one before 
                         moving on to the second:"""
-                    lm += inspect.cleandoc(prompt_str)
+                    lm += cleandoc(prompt_str)
 
                 with user():
                     prompt_str = f"""(1) From your perspective as an expert in {domain_expertise}, think step by step 
@@ -344,7 +344,7 @@ def request_ivs(
                        Where factor_name is one of the items within the factor_names list. If a factor does not have a high 
                        likelihood of being an instrumental variable, then do not wrap the factor with any tags. Your step by step 
                        answer as an in {domain_expertise}:"""
-                    lm += inspect.cleandoc(prompt_str)
+                    lm += cleandoc(prompt_str)
                 with assistant():
                     lm += gen("output")
 

pywhyllm/suggesters/model_suggester.py

@@ -5,7 +5,7 @@
 from guidance import system, user, assistant, gen
 from ..helpers import RelationshipStrategy
 import re
-import inspect
+from inspect import cleandoc
 
 class ModelSuggester(ModelerProtocol):
     CONTEXT: str = """causal mechanisms"""
@@ -39,7 +39,7 @@ def suggest_domain_expertises(
                     such factors? Think about this in a step by step manner and recommend {n_experts} expertises and 
                     provide each one wrapped within the tags, <domain_expertise></domain_expertise>, along with the 
                     reasoning and explanation wrapped between the tags <explanation></explanation>."""
-                    lm += inspect.cleandoc(prompt_str)
+                    lm += cleandoc(prompt_str)
                 with assistant():
                     lm += gen("output")
 
@@ -82,7 +82,7 @@ def suggest_domain_experts(
                     about this in a step by step manner and recommend {n_experts} domain experts and provide each one 
                     wrapped within the tags, <domain_expert></domain_expert>, along with the reasoning and explanation 
                     wrapped between the tags <explanation></explanation>."""
-                    lm += inspect.cleandoc(prompt_str)
+                    lm += cleandoc(prompt_str)
                 with assistant():
                     lm += gen("output")
 
@@ -126,7 +126,7 @@ def suggest_stakeholders(
                 this in a step by step manner and recommend {n_stakeholders} stakeholders. Then provide each useful stakeholder 
                 wrapped within the tags, <stakeholder></stakeholder>, along with the reasoning and explanation wrapped between the tags
                 <explanation></explanation>."""
-                    lm += inspect.cleandoc(prompt_str)
+                    lm += cleandoc(prompt_str)
                 with assistant():
                     lm += gen("output")
 
@@ -212,7 +212,7 @@ def request_confounders(
                     analysis_context}. Where a causal model is a conceptual model that describes the causal mechanisms of a 
                     system. You
                     will do this by answering questions about cause and effect and using your domain knowledge in {domain_expertise}."""
-                    lm += inspect.cleandoc(prompt_str)
+                    lm += cleandoc(prompt_str)
                 with user():
                     prompt_str = f"""Follow the next two steps, and complete the first one before moving on to the second: (1) 
                                 From your perspective as an 
@@ -234,7 +234,7 @@ def request_confounders(
                 <confounding_factor>factor_name</confounding_factor> where 
                 factor_name is one of the items within the factor_names list. If a factor does not have a high likelihood of directly 
                 confounding, then do not wrap the factor with any tags."""
-                    lm += inspect.cleandoc(prompt_str)
+                    lm += cleandoc(prompt_str)
                 with assistant():
                     lm += gen("output")
 
@@ -311,7 +311,7 @@ def suggest_parents(
 {factor}, 
                             then do not wrap the factor with any tags. Your answer as an expert in 
 {domain_expertise}:"""
-                    lm += inspect.cleandoc(prompt_str)
+                    lm += cleandoc(prompt_str)
 
                 with assistant():
                     lm += gen("output")
@@ -383,7 +383,7 @@ def suggest_children(
                             factor_names list. If a factor does not have a high likelihood of directly influencing and causing the {
                     factor}, then do not wrap the factor with any tags. Your answer as an expert in 
                 {domain_expertise}:"""
-                    lm += inspect.cleandoc(prompt_str)
+                    lm += cleandoc(prompt_str)
                 with assistant():
                     lm += gen("output")
 
@@ -424,7 +424,7 @@ def suggest_pairwise_relationship(
                             mechanisms of a system. You will do this by by answering questions about cause and effect and using your 
                             domain 
                             knowledge as an expert in {domain_expertise}."""
-                    lm += inspect.cleandoc(prompt_str)
+                    lm += cleandoc(prompt_str)
                 with user():
                     prompt_str = f"""From your perspective as an expert in {domain_expertise}, which of the following is 
                                     most likely true? (A) {factor_a} affects {factor_b}; {factor_a} has a high likelihood of directly 
@@ -436,7 +436,7 @@ def suggest_pairwise_relationship(
                 you reach a conclusion, wrap your answer within the tags <answer></answer>. If you are done thinking, provide your 
                 answer wrapped within the tags <answer></answer>. e.g. <answer>A</answer>, <answer>B</answer>, or <answer>C</answer>. 
                 Your answer as an expert in {domain_expertise}:"""
-                    lm += inspect.cleandoc(prompt_str)
+                    lm += cleandoc(prompt_str)
 
                 with assistant():
                     lm += gen("output")

pywhyllm/suggesters/tuebingen_model_suggester.py

@@ -120,17 +120,17 @@ def _build_description_program(self, variable, use_context=False, ask_reference=
             query["user"] = cleandoc(user_prompt)
 
             if ask_reference:
-                query["user"] += " " # cleandoc removes leading whitespace, so need to add separately
                 user_prompt = f"""Then provide two research papers that support your description.
                 Let's think step-by-step to make sure that we have a proper and clear description. Then provide your final 
-                answer within the tags, <description></description>, and each research paper within the tags <reference></reference>."""
-                query["user"] += cleandoc(user_prompt)
+                answer within the tags, <description></description>, and each research paper within the tags <reference></reference>.""" 
+                # cleandoc removes leading whitespace, so need to add separately
+                query["user"] += " " + cleandoc(user_prompt)
 
             else:
-                query["user"] += "\n" # cleandoc removes leading whitespace, so need to add separately
                 user_prompt = f"""Let's think step-by-step to make sure that we have a proper and clear description. Then provide your final 
                 answer within the tags, <description></description>."""
-                query["user"] += cleandoc(user_prompt)
+                # cleandoc removes leading whitespace, so need to add separately
+                query["user"] += "\n" + cleandoc(user_prompt)
 
         else:
             sys_prompt = f"""You are a helpful assistant for writing concise and peer-reviewed descriptions. Your goal 
@@ -141,17 +141,16 @@ def _build_description_program(self, variable, use_context=False, ask_reference=
             query["user"] = cleandoc(user_prompt)
 
             if ask_reference:
-                query["user"] += "\n" # cleandoc removes leading whitespace, so need to add separately
                 user_prompt = f"""Then provide two research papers that support your description.
                         Let's think step-by-step to make sure that we have a proper and clear description. Then provide 
                         your final answer within the tags, <description></description>, and each research paper within the 
                         tags <paper></paper>."""
-                query["user"] += cleandoc(user_prompt)
+                # cleandoc removes leading whitespace, so need to add separately
+                query["user"] += "\n" + cleandoc(user_prompt)
             else:
-                query["user"] += "\n" # cleandoc removes leading whitespace, so need to add separately
                 user_prompt = f"""Let's think step-by-step to make sure that we have a proper and clear description. Then provide 
                         your final answer within the tags, <description></description>."""
-                query["user"] += cleandoc(user_prompt)
+                query["user"] += "\n" + cleandoc(user_prompt)
         return query
 
     def suggest_relationship(
@@ -258,10 +257,10 @@ def _build_relationship_program(
                 if use_description:
                     user_prompt = f"""can changing {variable_a}, where {description_a}, change {variable_b}, where 
     {description_b}? Answer Yes or No."""
-                    query["user"] += cleandoc(user_prompt)
+                    query["user"] = cleandoc(user_prompt)
                 else:
                     user_prompt = f"""can changing {variable_a} change {variable_b}? Answer Yes or No."""
-                    query["user"] += cleandoc(user_prompt)
+                    query["user"] = cleandoc(user_prompt)
 
                 if ask_reference:
                     user_prompt = f"""At each step, each expert include a reference to a research paper that supports