Can I add prompt to create_pandas_dataframe_agent? is that possible?

[mraguth]

I like to create a prompt template and attach to create_pandas_dataframe_agent. Is that possible? sample code is below...

from langchain_openai import ChatOpenAI
from langchain_experimental.agents import create_pandas_dataframe_agent
import pandas as pd

df = pd.read_csv("titanic.csv")
llm = ChatOpenAI(model="gpt-3.5-turbo", temperature=0)
agent_executor = create_pandas_dataframe_agent(
    llm,
    df,
    agent_type="tool-calling",
    verbose=True
)

[dosubot[bot]]

Yes, you can add a prompt template to the create_pandas_dataframe_agent function. In the sample code you provided, you can define a prompt template and integrate it into the agent creation process. Here's an example of how you can do this:

from langchain_openai import ChatOpenAI
from langchain_experimental.agents import create_pandas_dataframe_agent
from langchain.prompts import ChatPromptTemplate, MessagesPlaceholder
import pandas as pd

df = pd.read_csv("titanic.csv")
llm = ChatOpenAI(model="gpt-3.5-turbo", temperature=0)

# Define your prompt template
TEMPLATE = """You are working with a pandas dataframe in Python. The name of the dataframe is `df`.
It is important to understand the attributes of the dataframe before working with it. This is the result of running `df.head().to_markdown()`

<df>
{dhead}
</df>

You are not meant to use only these rows to answer questions - they are meant as a way of telling you about the shape and schema of the dataframe.
You also do not have use only the information here to answer questions - you can run intermediate queries to do exploratory data analysis to give you more information as needed.

For example:

<question>How old is Jane?</question>
<logic>Use `person_name_search` since you can use the query `Jane`</logic>
<question>Who has id 320</question>
<logic>Use `python_repl` since even though the question is about a person, you don't know their name so you can't include it.</logic>
"""  # noqa: E501

# Format the template with the dataframe's head
template = TEMPLATE.format(dhead=df.head().to_markdown())

# Create a ChatPromptTemplate
prompt = ChatPromptTemplate.from_messages(
    [
        ("system", template),
        MessagesPlaceholder(variable_name="agent_scratchpad"),
        ("human", "{input}"),
    ]
)

# Create the agent with the prompt
agent_executor = create_pandas_dataframe_agent(
    llm,
    df,
    agent_type="tool-calling",
    prompt=prompt,
    verbose=True
)

This code shows how to define a prompt template and attach it to the create_pandas_dataframe_agent function ^[1].

To continue talking to Dosu, mention @dosu.