One of the most confusing things in Pandas is this:

df["region"]                 # column selection
df[df["region"] == "South"]  # row filtering

👉 Same df[...] syntax — but completely different behavior.
So how does Pandas know what to do?

---

🔷 Core Idea: df[...] Changes Behavior Based on Input Type

Pandas does type-based decision making.

👉 It doesn’t guess.
👉 It checks what you passed inside [].

---

🔹 Case 1: Column Selection

df["region"]

What happens internally:

df.__getitem__("region")

Pandas sees:

• Input = string

So it decides:

“You want a column”

✅ Output:

• A Series (the "region" column)

---

🔹 Case 2: Boolean Filtering

df[df["region"] == "South"]

Step 1: Inner expression runs first

df["region"] == "South"

👉 Produces:

0    False
1     True
2    False
3     True
dtype: bool

👉 This is a Boolean Series (mask)

---

Step 2: Outer df[...] receives the mask

df[BooleanSeries]

Internally:

df.__getitem__(BooleanSeries)

---

Pandas checks:

• Is it a Series/array? ✅
• Is dtype boolean? ✅
• Length matches rows? ✅

👉 Decision:

“This is a Boolean mask → filter rows”

---

✅ Result:

• Only rows where value is True are kept

---

🔷 Why df["region"] Does NOT Trigger Filtering

Because:

df["region"]

👉 Returns actual data, not True/False values

Example:

["North", "South", "East"]

❌ No Boolean values → no filtering
✔ Just column access

---

🔷 Side-by-Side Comparison

Expression	Input Type	Pandas Action	Output
df["region"]	string	Select column	Series
df[["region", "price"]]	list of strings	Select columns	DataFrame
df[mask]	Boolean Series	Filter rows	DataFrame

---

🔷 Mental Model (Very Important)

Think of df[...] like this:

def getitem(key):
    if key is string:
        return column

    elif key is list of column names:
        return multiple columns

    elif key is Boolean Series:
        return filtered rows

👉 Same syntax, different behavior based on type of input

---

🔷 Key Insight

Pandas does NOT understand your condition (== "South")
It only sees the result (True/False values)

---

🔷 Bonus Examples

Multiple columns:

df[["region", "price"]]

👉 Input = list → returns selected columns

---

Boolean list (mask):

df[[True, False, True, False]]

👉 If length matches → filters rows

---

🔥 Final Takeaway

• df[...] is polymorphic (same syntax, different meanings)
• Behavior depends on input type
• Boolean filtering works only when:
  • Values are True/False
  • Length matches DataFrame rows

---

🚀 Why This Matters

Understanding this helps you:

• Debug Pandas errors faster
• Avoid incorrect filtering
• Write cleaner data analysis code
• Perform well in interviews