Series indexing and selection in Data Analysis Python - Time & Space Complexity

Time Complexity: Series indexing and selection

O(1) for single item, O(k) for slice of size k

Understanding Time Complexity

When working with a Series, we often pick values by their position or label. Understanding how long this takes helps us write faster code.

We want to know: How does the time to get items change as the Series grows?

Scenario Under Consideration

Analyze the time complexity of the following code snippet.

import pandas as pd

s = pd.Series(range(1000))

# Select a single value by position
value = s.iloc[500]

# Select multiple values by label
subset = s.loc[100:200]

This code creates a Series and selects values by position and by label range.

Identify Repeating Operations

Identify the loops, recursion, array traversals that repeat.

• Primary operation: Accessing elements by index or label.
• How many times: Single access is once; slicing accesses multiple elements equal to slice size.

How Execution Grows With Input

Getting one item by position takes about the same time no matter the Series size.

Getting a slice of items takes time roughly proportional to how many items are in the slice.

Input Size (n)	Approx. Operations for single item	Approx. Operations for slice of size k
10	1	5
100	1	20
1000	1	50

Pattern observation: Single item access stays about the same; slice access grows with slice size.

Final Time Complexity

Time Complexity: O(1) for single item, O(k) for slice of size k

This means picking one item is very fast no matter how big the Series is, but picking many items takes longer as you ask for more.

Common Mistake

[X] Wrong: "Selecting any item from a Series always takes longer if the Series is bigger."

[OK] Correct: Single item access uses direct indexing, so it stays fast even if the Series grows large.

Interview Connect

Knowing how fast you can get data from a Series helps you choose the right method and write code that runs smoothly in real projects.

Self-Check

"What if we changed from selecting by label slice to selecting by a list of random labels? How would the time complexity change?"

Practice

1. What does series['a'] do in a pandas Series with index labels?

easy

A. Returns a slice from start to label 'a'

B. Selects the value at label 'a' in the Series

C. Selects the value at position 0 in the Series

D. Raises an error because 'a' is not a position

Solution

1. Step 1: Understand label-based indexing

   Using series['a'] accesses the value at the index label 'a', not position.

2. Step 2: Differentiate label vs position

   Label 'a' is a named index, so it returns the value at that label.

3. Final Answer:

   Selects the value at label 'a' in the Series -> Option B

4. Quick Check:

   Label indexing = value at label [OK]

Hint: Use labels inside quotes to get values by label [OK]

Common Mistakes:

• Confusing label with position indexing
• Thinking 'a' is a position index
• Expecting a slice instead of single value

2. Which of these is the correct syntax to select multiple labels 'a' and 'c' from a Series named ser?

easy

A. ser.loc[['a', 'c']]

B. ser['a', 'c']

C. ser.loc['a', 'c']

D. ser.iloc[['a', 'c']]

Solution

1. Step 1: Understand selection of multiple labels

   To select multiple labels, use loc with a list of labels inside double brackets.

2. Step 2: Check syntax correctness

   ser.loc[['a', 'c']] is correct; distractors B and C use tuples (KeyError), D uses iloc which expects positions (TypeError).

3. Final Answer:

   ser.loc[['a', 'c']] -> Option A

4. Quick Check:

   Use loc with list for multiple labels [OK]

Hint: Use ser.loc[[labels]] to select multiple labels [OK]

Common Mistakes:

• Using single brackets for multiple labels
• Omitting loc when selecting multiple labels
• Passing labels as separate arguments instead of a list

3. Given the Series:

import pandas as pd
ser = pd.Series([10, 20, 30, 40], index=['a', 'b', 'c', 'd'])
print(ser['b':'d'])

What is the output?

medium

A. b 20
c 30
d 40
dtype: int64

B. b 20
c 30
dtype: int64

C. 20
30
40

D. 20
30

Solution

1. Step 1: Understand label slicing in Series

   Label slicing with ser['b':'d'] includes both start and end labels.

2. Step 2: Identify values for labels 'b' to 'd'

   Labels 'b', 'c', and 'd' correspond to values 20, 30, and 40 respectively.

3. Final Answer:

   b 20 c 30 d 40 dtype: int64 -> Option A

4. Quick Check:

   Label slice includes end label [OK]

Hint: Label slices include the end label [OK]

Common Mistakes:

• Excluding the end label in slice
• Confusing label slicing with position slicing
• Expecting only values without index labels

4. What is wrong with this code snippet?

import pandas as pd
ser = pd.Series([1, 2, 3], index=['x', 'y', 'z'])
print(ser[1:2])

Why might this cause confusion?

medium

A. It raises a KeyError because 1 and 3 are not labels

B. It returns a single value instead of a slice

C. It returns values for labels 'y' and 'z' inclusive

D. It slices by position, excluding index label 'z'

Solution

1. Step 1: Understand position slicing in Series

   Using ser[1:2] slices by position, not label, and excludes the end position.

2. Step 2: Identify slice range and output

   Position 1 (inclusive) to 2 (exclusive) corresponds to label 'y' (position 1), excluding position 2 ('z').

3. Final Answer:

   It slices by position, excluding index label 'z' -> Option D

4. Quick Check:

   Position slices exclude end position [OK]

Hint: Position slices exclude the end index [OK]

Common Mistakes:

• Expecting label slicing with integer slice
• Thinking slice includes end position
• Confusing KeyError with slicing behavior

5. You have a Series ser with index labels ['a', 'b', 'c', 'd', 'e'] and values [5, 10, 15, 20, 25]. How do you select all values greater than 10 using Series indexing and selection?

hard

A. ser[ser.index > 10]

B. ser.index[ser > 10]

C. ser[ser > 10]

D. ser.loc[ser.index > 10]

Solution

1. Step 1: Use boolean indexing on Series values

   Comparing ser > 10 creates a boolean mask for values greater than 10.

2. Step 2: Apply boolean mask to Series

   Using ser[ser > 10] selects values where the mask is True.

3. Final Answer:

   ser[ser > 10] -> Option C

4. Quick Check:

   Boolean mask on values selects correct subset [OK]

Hint: Use ser[ser > value] to filter by value condition [OK]

Common Mistakes:

• Trying to filter by index values instead of data values
• Using loc with boolean mask on index instead of values
• Confusing index labels with values in condition