Online Robotics Classes for Kids in Orchard Road, Singapore

1. Introduction: Ordering and Ranking Structured Data

Tables rarely arrive in the order you need for analysis. Pandas can sort by one or more columns and assign ranks so you can surface top performers, lowest costs, or tiered groups.

• Single-column sort — ascending or descending
• Multi-column sort — primary key, then secondary to break ties (independent ascending flags)
• Grouped ranking — e.g. dense ranks within each group

2. Sorting Single and Multiple Columns (sort_values)

Sorting is usually done on numeric columns. When many rows share the same value (e.g. Quantity = 10), a single-column sort leaves those ties in arbitrary order. Pass a second column so ties are ordered by that secondary field.

3. Hierarchical Sorting in Practice

Pass a list of columns to sort_values(). Example: sort by Quantity (descending), then by Sales (ascending) when Quantity ties:

import pandas as pd

df = pd.read_csv("retail_sales.csv")

sorted_df = df.sort_values(
    by=["Quantity", "Sales"],
    ascending=[False, True]
)

print(sorted_df[["Quantity", "Sales"]].head(10))

Highest quantities (e.g. 19) rise to the top; within that quantity, Sales runs from low to high (e.g. 0, then 227, 361, 368).

4. Ranking Data in Pandas

Ranking assigns relative positions. Group by a primary column, then rank a secondary column inside each group.

Dense ranking (example within Quantity = 19): highest Sales → rank 1, next → rank 2, and so on, without skipping ranks on ties.

df["Sales_Rank_Within_Qty"] = (
    df.groupby("Quantity")["Sales"]
    .rank(method="dense", ascending=False)
)

ranked_df = df.sort_values(
    by=["Quantity", "Sales_Rank_Within_Qty"],
    ascending=[False, True]
)

print(ranked_df[["Quantity", "Sales", "Sales_Rank_Within_Qty"]].head(10))

Each Quantity tier gets its own rank set; top sales inside that tier receive rank 1 when ascending=False.