Select exactly what you need

df["Score"] selects one column as a Series. df[["Score"]] passes a list of column names and returns a DataFrame, even with only one column. To select several columns, put their names in that inner list. The distinction matters when another operation expects a table.

df = students()
column = df["Score"]
table = df[["Score"]]
result = df[["Name", "Score"]]
print(type(column).__name__, column.shape)
print(type(table).__name__, table.shape)
print(result.head(3))

Series (12,)
DataFrame (12, 1)
    Name  Score
0  Amira     88
1  Bilal     72
2  Chloe     91

Watch for this. df["Name", "Score"] is not the same as df[["Name", "Score"]] for these ordinary columns.

df = students()
result = df[["Campus", "Program"]]
print(result.head(4))
print(result.shape)

What you should be able to explain. I can predict whether a column selection returns a Series or a DataFrame.

set_index("Student_ID") uses the student IDs as row labels and, by default, removes that field from the data columns. loc has the form table.loc[row_labels, column_labels]. A list keeps the selection table-shaped. at is convenient when you need a single value with a row label and column label.

df = students().set_index("Student_ID")
result = df.loc[["S001", "S003"], ["Name", "Score"]]
one_score = df.at["S003", "Score"]
print(result)
print("One score:", one_score)

             Name  Score
Student_ID              
S001        Amira     88
S003        Chloe     91
One score: 91

Watch for this. set_index does not automatically check that IDs are unique; uniqueness needs its own check.

df = students().set_index("Student_ID")
result = df.loc[["S002", "S004"], ["Name", "Campus"]]
one_score = df.at["S004", "Score"]
print(result)
print("One score:", one_score)

What you should be able to explain. I can read selected records and one value using meaningful labels.

iloc uses numerical positions for both rows and columns. Positions start at 0 even when the printed index labels are student IDs. A slice excludes its stop position. iat reads one value using a row position and a column position.

df = students()[["Name", "Campus", "Score"]]
result = df.iloc[1:4, [0, 2]]
one_score = df.iat[1, 2]
print(result)
print("One score:", one_score)

    Name  Score
1  Bilal     72
2  Chloe     91
3   Dana     65
One score: 72

Watch for this. iloc expects positions, so df.iloc["S001"] is not valid positional selection.

df = students()[["Name", "Campus", "Score"]]
result = df.iloc[:2, :]
one_score = df.iat[0, 2]
print(result)
print("One score:", one_score)

What you should be able to explain. I can select a rectangular portion of a table by position.

For this sorted, unique index, a loc label slice includes both its start and stop labels. An iloc position slice excludes its stop position. Compare the selected IDs instead of assuming the two expressions return the same rows. Label slicing works best when the index is appropriately sorted and its labels are understood.

df = students().set_index("Student_ID").sort_index()
by_label = df.loc["S002":"S004", ["Name", "Score"]]
by_position = df.iloc[1:3][["Name", "Score"]]
print("Label slice:")
print(by_label)
print("Position slice:")
print(by_position)

Label slice:
             Name  Score
Student_ID              
S002        Bilal     72
S003        Chloe     91
S004         Dana     65
Position slice:
             Name  Score
Student_ID              
S002        Bilal     72
S003        Chloe     91

Watch for this. Selecting from a selection is fine for reading; use one .loc assignment when changing the original table.

df = students().set_index("Student_ID").sort_index()
by_label = df.loc["S002":"S004", ["Name", "Score"]]
by_position = df.iloc[1:4][["Name", "Score"]]
same = by_label.equals(by_position)
print(by_position)
print("Same result:", same)

What you should be able to explain. I can explain why a label slice and a position slice can return different rows.