Sorting and index decisions

sort_values() orders whole rows by one or more column values. ascending=False puts larger values first. It returns a new table, so assign the result if you want to keep the order. The index labels travel with their rows and are not automatically renumbered.

df = students()
result = df.sort_values("Score", ascending=False).head(3)
print(result[["Student_ID", "Name", "Score"]])
print("Index labels:", result.index.tolist())

  Student_ID    Name  Score
5       S006  Fatima     95
2       S003   Chloe     91
9       S010   Jamal     89
Index labels: [5, 2, 9]

Watch for this. head(3) before sorting chooses the first three source rows, not the three highest scores.

df = students()
result = df.sort_values("Score", ascending=True).head(3)[["Student_ID", "Name", "Score"]]
print(result)

What you should be able to explain. I can sort a table and identify the top or bottom records.

Rows can tie on a sort column. With one column, kind="stable" preserves their existing order within a tie. To define a new tie rule, sort by several columns and give one ascending flag per column. Clear tie rules make results easier to reproduce and explain.

df = pd.DataFrame({
    "Name": ["Chloe", "Amira", "Bilal", "Dana"],
    "Score": [80, 80, 95, 80]
})
result = df.sort_values("Score", ascending=False, kind="stable")
print(result)

    Name  Score
2  Bilal     95
0  Chloe     80
1  Amira     80
3   Dana     80

Watch for this. Alphabetical tie order must be requested explicitly; a stable sort preserves source order instead.

df = pd.DataFrame({
    "Name": ["Chloe", "Amira", "Bilal", "Dana"],
    "Score": [80, 80, 95, 80]
})
result = df.sort_values(["Score", "Name"], ascending=[False, True])
print(result)

What you should be able to explain. I can make the order of tied records explicit.

An index labels rows; it is not automatically a unique ID. set_index() chooses the labels, and index.is_unique tells you whether any label repeats. reset_index() moves the current index back into a column and supplies new default row labels. Use reset_index(drop=True) only when you deliberately want to discard the current labels.

df = students()
indexed = df.set_index("Student_ID")
print("Unique index:", indexed.index.is_unique)
result = indexed.reset_index()
print(result[["Student_ID", "Name"]].head(3))

Unique index: True
  Student_ID   Name
0       S001  Amira
1       S002  Bilal
2       S003  Chloe

Watch for this. Having an index does not guarantee that its labels are unique or non-missing.

df = students()
indexed = df.set_index("Student_ID")
selected = indexed.loc[["S003", "S001"]]
result = selected.reset_index()
print(result[["Student_ID", "Name"]])

What you should be able to explain. I can change an index without accidentally discarding record IDs.

reindex() aligns a table to a specified set of labels and their requested order. Existing labels bring their rows with them; new labels create missing values. This differs from sorting because you supply the desired labels explicitly. Use isna() to inspect newly missing cells.

df = students().set_index("Student_ID")
result = df.reindex(["S003", "S001", "S999"])[["Name", "Score"]]
print(result)
print("Missing name:")
print(result["Name"].isna())

             Name  Score
Student_ID              
S003        Chloe   91.0
S001        Amira   88.0
S999          NaN    NaN
Missing name:
Student_ID
S003    False
S001    False
S999     True
Name: Name, dtype: bool

Watch for this. Reindexing can fail when source labels are duplicated; check uniqueness before using labels as identifiers.

df = students().set_index("Student_ID")
result = df.reindex(["S002", "S001", "S004"])[["Name", "Score"]]
print(result)
print("Missing names:", result["Name"].isna().sum())

What you should be able to explain. I can distinguish sorting existing rows from aligning to a requested index.