Hashmap Patterns in LeetCode Problems (Medium)

Group 1: Basic Frequency Counting & Existence Checks

This group introduces the most fundamental use of hash maps: counting element occurrences or checking for their existence. Problems start with simple counting and progress to using maps to track state across more complex data structures like a 2D grid.

#	Problem	Link	Concept	Description of Difficulty Increase
1	Find All Duplicates in an Array	Link	Frequency Counting	Base case: Use a hash map to count integer occurrences and identify those that appear twice.
2	Rabbits in Forest	Link	Frequency Counting	Adds a layer of logic on top of frequency counts. You must interpret the counts according to the problem's grouping rules.
3	Bulls and Cows	Link	Frequency Counting	Requires two passes or a single clever pass. Introduces managing two types of counts (bulls vs. cows) based on character and position.
4	K-diff Pairs in an Array	Link	Frequency Counting	Builds on counting by requiring a second pass over the map's keys to find pairs that satisfy a specific condition (difference of k).
5	Valid Sudoku	Link	Existence Check (Hash Set)	Applies the existence check pattern to a 2D grid. Requires managing 27 different sets (9 rows, 9 columns, 9 boxes) simultaneously.
6	Set Matrix Zeroes	Link	State Tracking (Hash Set)	Uses hash sets not for counting, but to store the rows and columns that need to be modified, separating the "finding" and "setting" phases.
7	Subdomain Visit Count	Link	Frequency Counting	Adds significant string parsing logic. You must first process the input strings into subdomains before applying the counting pattern.
8	Finding the Users Active Minutes	Link	Frequency Counting	Involves a nested data structure: a hash map where each value is a hash set, to count unique minutes per user.
9	Display Table of Food Orders in a Restaurant	Link	Frequency Counting	Requires managing nested counts (table → food item → count). Adds complexity of collecting all keys for table headers and sorting.

Group 2: Grouping by Canonical Representation

These problems use a hash map to group items. The key to the map is a “canonical” (standard) representation of an item, and the value is a list of all items that map to that key.

#	Problem	Link	Concept	Description of Difficulty Increase
10	Group Anagrams	Link	Grouping	Foundational grouping problem. The canonical key is the sorted version of a string.
11	Find and Replace Pattern	Link	Grouping / Pattern Matching	Introduces a more complex canonical key. You need to normalize each word into a numeric pattern (e.g., 'abb' → '122') to group them.
12	Groups of Special-Equivalent Strings	Link	Grouping	A variation on the canonical key concept. The key must represent the counts of characters at both even and odd indices separately.
13	Determine if Two Strings Are Close	Link	Frequency Map Comparison	Instead of grouping, you create two frequency maps and compare them based on two conditions: same keys and same frequency values.
14	Word Subsets	Link	Frequency Map Comparison	Requires creating a single "max frequency" map from one list of words and then comparing each word from another list against it.
15	Find Duplicate File in System	Link	Grouping	Adds string parsing to extract file content, which serves as the canonical key for grouping file paths.

Group 3: Frequency Counting with Sorting or Greedy Logic

Here, the hash map is the first step in a two-step process. After counting frequencies, you must apply a sorting or greedy algorithm to the counts to find the answer.

#	Problem	Link	Concept	Description of Difficulty Increase
16	Sort Characters by Frequency	Link	Frequency Count & Sort	Base case: Count character frequencies, then sort the characters based on those counts to build the result string.
17	Top K Frequent Elements	Link	Frequency Count & Heap/Sort	Similar to the previous, but introduces finding the "top K" elements, making a Min-Heap or Quickselect an efficient follow-up step.
18	Top K Frequent Words	Link	Frequency Count & Heap/Sort	A "boring variation" that adds complexity to the sorting logic: you must handle ties using lexicographical order.
19	Reduce Array Size to the Half	Link	Frequency Count & Greedy	Introduces a greedy approach. After counting, you sort the frequencies and greedily remove the largest counts to solve the problem.
20	Least Number of Unique Integers after K Removals	Link	Frequency Count & Greedy	Similar greedy logic, but now you remove the smallest counts first. Reinforces the pattern of operating on sorted frequencies.
21	Hand of Straights	Link	Frequency Count & Greedy	Requires more complex greedy logic. After counting, you must check for consecutive sequences, often by iterating through sorted keys.
22	Divide Array in Sets of K Consecutive Numbers	Link	Frequency Count & Greedy	A "boring variation" of Hand of Straights. The underlying logic is identical.
23	Reorganize String	Link	Frequency Count & Greedy Heap	The greedy logic becomes more complex, requiring a Max-Heap to always pick the most frequent character that is not adjacent to the last.
24	Distant Barcodes	Link	Frequency Count & Greedy Heap	A "boring variation" of Reorganize String with the same underlying greedy heap-based algorithm.

Group 4: Modulo Arithmetic & Pair/Tuple Counting

This set of problems combines frequency maps with modulo arithmetic or number theory to find pairs or combinations that satisfy a specific sum or divisibility property.

#	Problem	Link	Concept	Description of Difficulty Increase
25	Pairs of Songs with Total Durations Divisible by 60	Link	Hashing with Modulo	Foundational problem for this pattern. Use a map to store counts of `time % 60` to find complementary pairs.
26	Check if Array Pairs Are Divisible by k	Link	Hashing with Modulo	A "boring variation" of the previous problem, generalizing the logic from a fixed `60` to a variable `k`.
27	4Sum II	Link	Hashing Pairs (Two-Sum)	Applies the two-sum pattern on a larger scale. You pre-compute and store sums of pairs from two arrays in a map to find `-(c+d)`.
28	Number of Pairs of Interchangeable Rectangles	Link	Hashing with Ratios	Uses a map to count occurrences of a canonical key, which is the reduced fraction (ratio) of width to height.
29	Count Good Meals	Link	Hashing Pairs (Two-Sum)	A variation of Two-Sum where the target isn't fixed. For each number, you must check for complements for all possible powers of two.
30	Tuple with Same Product	Link	Hashing Pairs	Uses a hash map to count the frequency of products of pairs. The final calculation requires combinatorial logic (nC2).

Group 5: Sliding Window with Hash Maps

In these problems, a hash map is used to maintain the state of a “sliding window” as it moves across an array or string. This is a powerful and common pattern.

#	Problem	Link	Concept	Description of Difficulty Increase
31	Longest Substring Without Repeating Characters	Link	Sliding Window	The classic introduction to using a map in a sliding window to track characters and their last seen positions.
32	Fruit Into Baskets	Link	Sliding Window	A slight variation where the window's constraint is having at most 2 distinct character types (fruits).
33	Permutation in String	Link	Sliding Window (Fixed)	Introduces a fixed-size window. Requires two maps: one for the target string and one for the current window, which you compare.
34	Find All Anagrams in a String	Link	Sliding Window (Fixed)	A "boring variation" of the previous problem. Instead of returning true/false, you collect all starting indices of valid anagrams.
35	Maximum Erasure Value	Link	Sliding Window	Combines the unique-element window from "Longest Substring" with calculating a sum, rather than just the length of the window.
36	Longest Repeating Character Replacement	Link	Sliding Window	Introduces a more complex window condition: `window_length - max_frequency <= k`. Requires efficiently tracking the max frequency.
37	Find the Longest Substring Containing Vowels in Even Counts	Link	Sliding Window & Bitmask	Advanced pattern combining a hash map with bitmasking. The map stores `mask → first_index` to find subarrays with a target parity state.

Group 6: Prefix Sum with Hash Maps

This pattern is used to efficiently find subarrays or subsequences that sum to a specific target. A hash map stores prefix sums and their indices, allowing for O(1) lookups.

#	Problem	Link	Concept	Description of Difficulty Increase
38	Subarray Sum Equals K	Link	Prefix Sum	The canonical problem for this pattern. You store counts of prefix sums in a map to find `current_sum - k`.
39	Binary Subarrays with Sum	Link	Prefix Sum	A "boring variation" of Subarray Sum Equals K, cementing the core concept.
40	Count Number of Nice Subarrays	Link	Prefix Sum	A re-framing of the same problem. You must first transform the input (odd=1, even=0) to see it as a standard subarray sum problem.
41	Contiguous Array	Link	Prefix Sum	Another re-framing. By converting 0s to -1s, the problem becomes finding the longest subarray with a sum of 0.
42	Subarray Sums Divisible by K	Link	Prefix Sum with Modulo	Combines the prefix sum pattern with modulo arithmetic. The map stores counts of `prefix_sum % k`.
43	Continuous Subarray Sum	Link	Prefix Sum with Modulo	A variation of the previous problem that requires checking for a subarray length of at least 2.
44	Minimum Operations to Reduce X to Zero	Link	Prefix Sum	Requires an insightful re-framing: the problem is equivalent to finding the longest middle subarray that sums to `total_sum - x`.
45	Remove Zero Sum Consecutive Nodes from Linked List	Link	Prefix Sum on Linked List	Applies the prefix sum pattern to a linked list. The map stores `prefix_sum → node`, requiring pointer manipulation to remove sublists.

Group 7: Graph/Tree Traversal with Hashing

In graph and tree problems, hash maps are essential for tracking visited nodes (to avoid cycles) or for mapping old nodes to new nodes when creating a deep copy.

#	Problem	Link	Concept	Description of Difficulty Increase
46	Employee Importance	Link	Graph Traversal (DFS/BFS)	Base case: Use a hash map to build an adjacency list (`id → Employee`) for easy lookups, then perform a simple traversal (DFS or BFS) to sum values.
47	Linked List Cycle II	Link	Traversal & Visited Set	Uses a hash set to keep track of visited nodes in a linked list. The first repeated node found is the start of the cycle.
48	Clone Graph	Link	Graph Traversal & Cloning	The canonical graph cloning problem. A hash map (`old_node → new_node`) is crucial for tracking visited/cloned nodes to avoid infinite loops and re-creations.
49	Copy List with Random Pointer	Link	Graph Traversal & Cloning	A "boring variation" of Clone Graph applied to a linked list with an extra `random` pointer, reinforcing the `old_node → new_node` map pattern.
50	Open the Lock	Link	BFS & Visited Set	A classic state-space search using BFS. A hash set is essential to store `visited` states (lock combinations) to prevent redundant exploration.
51	Minimum Genetic Mutation	Link	BFS & Visited Set	A "boring variation" of Open the Lock. The states are gene strings instead of numbers, but the BFS with a visited set pattern is identical.
52	Construct Binary Tree from Preorder and Inorder Traversal	Link	Recursion & Index Lookup	A hash map is used as a critical optimization to store `value → index` for the inorder traversal, allowing O(1) lookup of the root's position.
53	Construct Binary Tree from Inorder and Postorder Traversal	Link	Recursion & Index Lookup	A "boring variation" of the previous problem, using the same inorder map but with postorder traversal logic.
54	Restore the Array from Adjacent Pairs	Link	Graph Traversal	Requires building a graph (adjacency list via hash map) from pairs, finding an endpoint (a node with one neighbor), and then traversing to reconstruct.

Group 8: Dynamic Programming with Hash Map Memoization

In these problems, a hash map is used as a “memo” to store the results of previously computed subproblems, which is the core of top-down dynamic programming.

#	Problem	Link	Concept	Description of Difficulty Increase
55	Word Break	Link	DP with Memo	Classic DP problem. A hash set provides O(1) dictionary lookups, and a hash map can be used for memoization (`index → boolean`) to store subproblem results.
56	Delete and Earn	Link	DP with Memo	The first step is using a hash map to aggregate points for each number. The problem then becomes a "House Robber" variation on the sorted keys of the map.
57	Longest Arithmetic Subsequence of Given Difference	Link	DP with Memo	Introduces DP where the map state is `value → length`. For each `num`, the length of the sequence is `1 + map[num - difference]`.
58	Longest String Chain	Link	DP with Memo	The DP state is `word → longest_chain_length`. Requires sorting the words by length and building the solution by checking predecessors of each word.
59	Binary Trees with Factors	Link	DP with Memo	The DP state is `number → count_of_trees`. For each number, you iterate through its factors to calculate the total combinations.
60	Longest Arithmetic Subsequence	Link	DP with Memo	The DP state is more complex, requiring a nested map: `dp[index][difference]`. A map of maps (`map<int, map<int, int>>`) is a natural fit for this.

Group 9: Custom Data Structure Design

This group contains problems that require you to design and implement a class with specific methods and time complexities. A hash map is almost always the core component.

#	Problem	Link	Concept	Description of Difficulty Increase
61	Encode and Decode TinyURL	Link	Design	The simplest design problem. Requires two hash maps for forward and reverse lookups (`long→short` and `short→long`).
62	Insert Delete GetRandom O(1)	Link	Design	A classic design problem combining a hash map (`value → index`) with an array for O(1) average time complexity on all operations.
63	Time-Based Key-Value Store	Link	Design	Introduces combining a hash map with another structure. The map's value is a sorted list of `(timestamp, value)` pairs, requiring binary search for `get`.
64	Design Underground System	Link	Design	Requires multiple hash maps to track different states: one for active check-ins and another to aggregate travel times between stations.
65	Snapshot Array	Link	Design	A clever design for versioning. Each array index contains a map (`snap_id → value`) or a sorted list to allow efficient historical lookups.
66	LRU Cache	Link	Design	A canonical design problem requiring a combination of a hash map (for O(1) lookups) and a doubly linked list (to manage recency).
67	Design Twitter	Link	Design	A more complex design involving multiple hash maps (`userId → follows`, `userId → tweets`) and merging sorted lists or using a heap to generate the feed.
68	Stock Price Fluctuation	Link	Design	Combines a hash map for current prices with a sorted data structure (like a sorted map or two heaps) to efficiently track min and max prices.
69	Detect Squares	Link	Design	Requires storing point frequencies in a map. The `count` method involves iterating through existing points to find valid squares.

Group 10: Clever Hashing & Miscellaneous Problems

This final group includes problems where the hash map is used in a particularly clever way, as a minor utility in a larger algorithm, or for complex state tracking.

#	Problem	Link	Concept	Description of Difficulty Increase
70	Longest Consecutive Sequence	Link	Clever Hashing (Set)	A classic problem solved with a clever trick: use a hash set for O(1) lookups and only start counting a sequence from its true starting number.
71	Letter Combinations of a Phone Number	Link	Backtracking & Lookup	The hash map is used as a simple utility (a lookup table for `digit → letters`). The core logic is backtracking, not hashing.
72	Brick Wall	Link	Clever Hashing	The key insight is to use a hash map to count the frequency of "gaps" between bricks. The answer is `total_rows - max_gap_frequency`.
73	Fraction to Recurring Decimal	Link	Simulation & State	A complex simulation of long division. The hash map is critical for detecting a cycle by storing `remainder → position`.
74	Prison Cells After N Days	Link	Simulation & Cycle Detection	Since the state space is small, cycles must occur. A hash map (`state → day`) is used to detect the cycle and calculate the final state.
75	Smallest Integer Divisible by K	Link	Math & Cycle Detection	Similar to the above, this problem uses a hash set to detect when a `remainder % k` repeats, which indicates a non-terminating sequence.
76	Partition Labels	Link	Greedy & Pre-computation	A two-pass greedy solution. The first pass uses a hash map to pre-compute the last occurrence index of every character.
77	Task Scheduler	Link	Greedy & Math	The hash map is used for initial frequency counting. The core logic is a mathematical or greedy approach to calculate idle time.