Is there a possibility to do a regression/correlation analysis of ESG KPIs and the ESG Score using A

grae.tejada
grae.tejada
edited 7:14AM in Workspace SDK

Client wants to pull ESG KPIs and the ESG Score through API and do a regression analysis to see if the values affect each other.

Tagged:

Answers

  • jonathan.legrand
    jonathan.legrand

    Hi @grae.tejada, Please note that this Forum is for technical LSEG API questions only. For answers about content, such as the data item names (e.g.: what is the name of a "ESG KPI" field in the LDL), please use the DIB; failing that, please reach out to my.refinitiv.com.


    I looks into the DIB myself and tested out an answer for you with the fields ["TR.TRESGScore","TR.TRESGCControversiesScore"], and came up with:



    df = ld.get_history(
        universe="LSEG.L",
        fields=["TR.TRESGScore","TR.TRESGCControversiesScore"],
        start="2020-01-01",
        end="2024-09-01")

    df.corr()


    This is the output; is it what you're after?


    1726131098673.png



    You can go further in your analysis with:


    # Assuming df contains the necessary columns
    X = df[['TR.TRESGCControversiesScore']]
    y = df['TR.TRESGScore']


    # Handle any missing values if necessary
    X = X.fillna(X.mean())
    y = y.fillna(y.mean())


    # Initialize and fit the regression model
    model = LinearRegression()
    model.fit(X, y)


    # Print the coefficients
    print("Coefficient:", model.coef_[0])
    print("Intercept:", model.intercept_)


    # Predict ESG Score based on Controversies Score
    y_pred = model.predict(X)


    from sklearn.metrics import r2_score


    r2 = r2_score(y, y_pred)
    print("R-squared:", r2)



    Don't hesitate to let me know what Data Items you would prefer using, and I can entre them in the code above for you. Alternatively, I would suggest you do so yourself using Codebook.

Categories