Every analysis starts the same way: load, then look.
Python
import pandas as pd
df = pd.read_csv("orders.csv", parse_dates=["order_date"])
df.head() # first rows — is the shape what I expected?
df.dtypes # did types survive the trip?
df.describe() # ranges, means, and the outliers hiding in max
Two minutes of looking saves an hour of debugging a "weird result" that was really a string column pretending to be numbers.