The head of growth sends you orders.csv and asks for “a quick revenue number”. The fastest way to give a wrong answer is to load the file and sum a column straight away. Every experienced analyst does the same thing first: look at the data. Not the whole thing, a few deliberate checks that take thirty seconds and catch most problems before they reach a slide.
import pandas as pd\n\norders = pd.read_csv(\"orders.csv\")\nprint(orders.shape)\nprint(orders.head().to_string())read_csv reads the header row as column names, guesses each column’s type, and returns a DataFrame. head() shows the first five rows (head(10) shows ten, tail() the last ones, sample(5) five at random, which is better at surfacing oddities than the first rows). .to_string() prints every column instead of hiding the middle ones behind ....
2,066 rows and 7 columns. Before going further, say out loud what you expect: one row per order, an order id, the customer, a date, a status, a channel, an optional coupon and a shipping fee. The checks below confirm or refute each expectation.
\nimport pandas as pd\n\norders = pd.read_csv(\"orders.csv\")\norders.info()info() prints by itself, no print() needed. Read it column by column. Non-Null Count tells you how many rows have a value: coupon_code has 388 of 2,066, so most orders used no coupon. Dtype tells you what pandas made of each column: the ids are int64, the fee is float64, and everything else is object, meaning text.
That includes order_date. Stored as text, dates sort correctly only because this file happens to write them year first, and they cannot answer “which month?”. Ask read_csv to parse them as they load:
import pandas as pd\n\norders = pd.read_csv(\"orders.csv\", parse_dates=[\"order_date\"])\nprint(orders.dtypes)\nprint(orders[\"order_date\"].min(), \"to\", orders[\"order_date\"].max())Now the column is datetime64[ns], and min() and max() confirm the data runs from 2 January 2024 to 30 December 2025. Checking the date range of every new file is a habit worth forming: a missing month or a stray date in 1970 shows up immediately.
import pandas as pd\n\norders = pd.read_csv(\"orders.csv\")\nprint(orders.describe().round(2))describe() gives count, mean, standard deviation, minimum, quartiles and maximum for every numeric column. Shipping fees run from 0 to 9.99 with a median of 4.99: plausible. But it also dutifully reports a mean customer_id of 300.47, which means nothing. Ids are labels that happen to be numbers. Read describe() only for columns where arithmetic makes sense, and use it to spot impossible values: negative quantities, prices of zero, a maximum a hundred times the median.
import pandas as pd\n\norders = pd.read_csv(\"orders.csv\")\nprint(orders[\"status\"].value_counts())\nprint(orders[\"coupon_code\"].value_counts(dropna=False))\nprint(orders[\"channel\"].value_counts(normalize=True).round(3))value_counts() lists every distinct value with its count, most common first. It is how you find typos ("Web" next to "web"), unexpected categories, and how big each group is. dropna=False includes the missing values as their own row, and normalize=True gives shares instead of counts: 55% of orders came through the web.
Notice that 105 orders were cancelled. Whether revenue should include them is a business decision, not a coding one, and you will make it explicitly in Section 4. Inspection is where those decisions first become visible.
\nFinally, turn your expectations into one-line checks:
\nimport pandas as pd\n\norders = pd.read_csv(\"orders.csv\")\ncustomers = pd.read_csv(\"customers.csv\")\nprint(orders[\"order_id\"].is_unique) # one row per order?\nprint(orders[\"customer_id\"].nunique(), \"customers ordered\")\nprint(len(customers), \"customers in total\")\nprint(customers[\"referred_by\"].dtype)The order id is unique, so one row really is one order. 515 of 600 customers have ordered at least once; the other 85 signed up and never bought, which matters later when you compute “revenue per customer”. And referred_by loaded as float64 even though it holds customer ids: most customers were not referred, the blanks became NaN, and in classic pandas NaN is a float, so the whole column became one.
In the exercise you run the routine on customers.csv. Next you learn to pull out exactly the rows and columns you need with .loc and .iloc.
pd.read_csv("file.csv") loads a table; parse_dates=[...] turns date columns into real datetimes as they load.","Run the same routine on every new file - shape, head(), info(), describe(), then value_counts() on the text columns.","info() shows each column’s dtype and non-null count, which reveals missing values and numbers stored as text.","An integer column that contains blanks loads as float64, because classic NaN is a float.","Check the facts you are about to rely on, such as an id being unique, with code like orders["order_id"].is_unique."],"further":[{"title":"pandas.read_csv","url":"https://pandas.pydata.org/docs/reference/api/pandas.read_csv.html"},{"title":"pandas.DataFrame.info","url":"https://pandas.pydata.org/docs/reference/api/pandas.DataFrame.info.html"},{"title":"pandas.Series.value_counts","url":"https://pandas.pydata.org/docs/reference/api/pandas.Series.value_counts.html"}],"quiz":[{"q":"info() reports coupon_code 388 non-null object for a table of 2,066 rows. What does that tell you?
nunique() counts distinct values."},{"text":"388 rows failed to load and were dropped.","why":"read_csv keeps every row; the other 1,678 rows are present with a missing coupon."},{"text":"1,678 orders have no coupon code, and the rest hold text.","why":"Correct. 2,066 minus 388 are missing, and object means the stored values are strings (or a mix)."},{"text":"The column should be converted to numbers.","why":"Coupon codes like WELCOME10 are labels; object is the right dtype for them in pandas 2."}],"answer":2},{"q":"describe() shows mean 300.47 for customer_id. How should you read that?
describe() summarises every numeric column, including ids. Read it only for columns where arithmetic makes sense."},{"text":"Customer 300 is the most typical customer.","why":"An average of labels does not identify a typical anything. You would need a different question and a different calculation."}],"answer":1},{"q":"In customers.csv, referred_by holds customer ids but loads as float64. Why?
read_csv always loads id columns as floats.","why":"customer_id in the same file loads as int64. The difference is missing values."},{"text":"The column has blanks, which load as NaN, and classic NaN only exists in float columns.","why":"Correct. 503 customers were not referred, so pandas uses float64 to hold NaN. Lesson 3.1 shows the nullable Int64 type that avoids this."}],"answer":2}],"exercise":{"kind":"python","files":["customers.csv"],"starter":"import pandas as pd\n\n# Load customers.csv with signup_date parsed as a date\ncustomers = None\n\nn_rows = None # number of customers (rows)\nn_countries = None # how many distinct countries\ntop_country = None # the country with the most customers\nmissing_referrals = None # how many rows have no referred_by value\nfirst_signup = None # the earliest signup_date\n","solution":"import pandas as pd\n\ncustomers = pd.read_csv(\"customers.csv\", parse_dates=[\"signup_date\"])\ncustomers.info()\n\nn_rows = len(customers)\nn_countries = customers[\"country\"].nunique()\ntop_country = customers[\"country\"].value_counts().idxmax()\nmissing_referrals = customers[\"referred_by\"].isna().sum()\nfirst_signup = customers[\"signup_date\"].min()\n","checks":[{"id":"loaded","test":"isinstance(customers, pd.DataFrame) and customers.shape == (600, 9)"},{"id":"dates-parsed","test":"pd.api.types.is_datetime64_any_dtype(customers['signup_date'])"},{"id":"counts","test":"n_rows == 600 and n_countries == 8"},{"id":"top-country","test":"top_country == 'United States'"},{"id":"missing","test":"int(missing_referrals) == 503"},{"id":"first-signup","test":"str(pd.Timestamp(first_signup).date()) == '2024-01-01'"}],"title":"Inspect the customers table","prompt":"Load customers.csv into customers, parsing signup_date as a date while loading. Then answer five questions with code: n_rows (number of rows), n_countries (number of distinct countries), top_country (the country with the most customers), missing_referrals (how many rows have no referred_by), and first_signup (the earliest signup date).
pd.read_csv("customers.csv", parse_dates=["signup_date"]) loads and converts in one step.","nunique() counts distinct values; value_counts() sorts by count, so its first label is the most common.","customers["referred_by"].isna().sum() counts blanks: True counts as 1. customers["signup_date"].min() gives the earliest date."],"explanation":"len(df) counts rows, and nunique() confirms Cartwheel sells to exactly eight countries, so there are no spelling variants hiding in this clean table (the messy export is a different story). value_counts().idxmax() returns the label with the highest count: the United States, with 104 customers. Summing a boolean Series counts its True values, which is the standard way to count blanks. Because signup_date was parsed, min() compares real dates rather than text and returns a Timestamp, pandas’ date-and-time type.
customers has 600 rows and 9 columns","dates-parsed":"signup_date is a datetime column","counts":"600 rows and 8 countries","top-country":"The top country is the United States","missing":"503 customers have no referral","first-signup":"The first signup was on 2024-01-01"}},"words":725,"readMinutes":3}