transform_pipeline_step.py

# Compute GDP by using GDP per capita and the Population columns
df_final["GDP"] = df_final["GDP per capita"] * df_final["Population"]

# Remove % sign of Population under 20 years old column and convert it to be of type float
def transform_col(col_val):
    try: 
        return float(col_val.replace(" %", ""))
    except: # value is NaN
        return col_val

df_final["Population under 20 years old in %"] = df_final["Population under20 years old"].apply(transform_col)
df_final = df_final.drop(columns=["Population under20 years old"])