Level 1 · Slice a week of orders

A list holds values in order

A list is an ordered collection written with square brackets:

daily_orders = [42, 38, 51, 47, 60, 73, 68]
cart = ["Oak Side Table", "Linen Throw Pillow", "Brass Floor Lamp"]

Lists can hold any kind of value, and they keep the order you put things in.

Indexing starts at zero

You get a single item with its index in square brackets. The first item is index 0, not 1:

daily_orders[0]    # 42   Monday
daily_orders[1]    # 38   Tuesday
daily_orders[6]    # 68   Sunday
daily_orders[7]    # IndexError: list index out of range

Negative indexes count from the end, which is handy when you don’t know the length:

daily_orders[-1]   # 68   the last item
daily_orders[-2]   # 73   second to last

Slicing takes a range

A slice list[start:stop] returns a new list from start up to, but not including, stop. Leave out either end to go to the edge:

daily_orders[0:5]   # [42, 38, 51, 47, 60]   Monday to Friday
daily_orders[:5]    # same thing
daily_orders[5:]    # [73, 68]               the weekend
daily_orders[-3:]   # [60, 73, 68]           the last three days
daily_orders[::2]   # [42, 51, 60, 68]       every second day

The “up to but not including” rule feels odd at first, but it has a nice property: [:5] gives exactly 5 items, and [:5] plus [5:] together are the whole list with nothing missing or repeated.

Built-ins that work on lists

len(daily_orders)    # 7     number of items
sum(daily_orders)    # 379   total orders for the week
max(daily_orders)    # 73    busiest day
min(daily_orders)    # 38    quietest day
sum(daily_orders) / len(daily_orders)   # 54.14...  average per day

These are everyday analyst moves. “What was the busiest day?” is max; “which day was it?” is daily_orders.index(max(daily_orders)), which returns 5, meaning Saturday.

Changing a list

Unlike strings, lists can be changed in place:

daily_orders.append(55)      # add to the end
daily_orders[0] = 44         # correct Monday's figure

Be careful: append changes the list itself and returns None, so daily_orders = daily_orders.append(55) leaves you with None instead of a list.

Lists can hold anything

A list doesn’t have to hold numbers. It can hold strings, booleans, or even other lists, and the values don’t all need the same type, although in analysis work they almost always do. A list of product names, a list of order IDs to check, a list of column names to keep: you’ll build lists like these constantly, and every one of them supports the same indexing, slicing and len().

Common mistakes

• Off-by-one errors: daily_orders[1:5] starts at Tuesday because index 1 is the second item.
• Expecting daily_orders[5] to be Friday. It’s Saturday; Friday is index 4.
• Overwriting a list with the return value of append, sort or other in-place methods.

If you know SQL

Slicing a list is a bit like LIMIT and OFFSET on rows that already have a fixed order. You’ll see the same idea again in pandas with .head() and .iloc.