Comparing two columns in dataframes to get the different ones

pd_merge_conditional.py

import pandas as pd
import numpy as np

df1 = pd.DataFrame(columns=["product_id", "PMID"])
df2 = pd.DataFrame(columns=["product_id", "PMID"])

# El generado de ahora
df1.loc[0] = [1, 111]
df1.loc[1] = [2, 222]
df1.loc[2] = [3, 333]

# el nuevo
df2.loc[0] = [2, 888]
df2.loc[1] = [3, 333]

# Nuevo
# "1", "111"
# "2", "222"

# Old
# "3", "333"
df_merge = df1.merge(df2, how="left", on=["product_id"])
old = df_merge[df_merge.PMID_x==df_merge.PMID_y]
new = df_merge[df_merge.PMID_x!=df_merge.PMID_y]
new = new.rename(columns={"PMID_x": "PMID"})
old = old.rename(columns={"PMID_x": "PMID"})
print(new[["product_id", "PMID"]])
print(old[["product_id", "PMID"]])