Say I have an array like this
string [] fruits = {"watermelon","apple","apple","kiwi","pear","banana"};
Is there an built in function that allows me to query all the index of "apple" ?
For example,
fruits.FindAllIndex("apple");
will return an array of 1 and 2
If there is not, how should I implement it?
Thanks!
LINQ version:
var indexes = fruits.Select((value, index) => new { value, index })
.Where(x => x.value == "apple")
.Select(x => x.index)
.ToList();
Non-LINQ version, using Array<T>.IndexOf() static method:
var indexes = new List<int>();
var lastIndex = 0;
while ((lastIndex = Array.IndexOf(fruits, "apple", lastIndex)) != -1)
{
indexes.Add(lastIndex);
lastIndex++;
}
One way would be to write like this:
var indices = fruits
.Select ((f, i) => new {f, i})
.Where (x => x.f == "apple")
.Select (x => x.i);
Or the traditional way:
var indices = new List<int>();
for (int i = 0; i < fruits.Length; i++)
if(fruits[i] == "apple")
indices.Add(i);
Pretty easy with an extension method.
var fruits = new[] { "watermelon","apple","apple","kiwi","pear","banana" };
var indexes = fruits.FindAllIndexes("apple");
public static class Extensions
{
public static int[] FindAllIndexes(this string[] array, string search) => array
.Select((x, i) => (x, i))
.Where(value => value.x == search)
.Select(value => value.i)
.ToArray();
}
This isn't a question I feel I have the vocabulary to properly express, but I have two collections of the same anonymous type (lets call it 'a.)
'a is defined as new {string Name, int Count}
One of these collections of 'a we shall call requirements.
One of these collections of 'a we shall call candidates.
Given these collections, I want to determine if the following assertions hold.
If there exists some element in requirements r such that r.Count == 0, each element in candidates c such that r.Name == c.Name must satisfy c.Count == 0. There must exist one such element in candidates for each such element in requirements.
For each element of requirements r where r.Count > 0, there must be some subset of elements in candidates c such that c₁.Name, c₂.Name, ..., cₓ.Name == r.Name and that c₁ + ... + cₓ >= r.Count. Each element of candidates used to satisfy this rule for some element in requirements may not be used for another element in requirements.
An example of this would be that given
requirements = {{"A",0}, {"B", 0}, {"C", 9}}
candidates = {{"B", 0}, {"C", 1}, {"A",0}, {"D", 2}, {"C", 4}, {"C", 4}}
That this query would be satisfied.
r={"A", 0} and r={"B", 0} would be satisfied according to rule #1 against c={"A", 0} and c={"B", 0}
-and-
r={"C", 9) is satisfied according to rule #2 by the group gc on collections c.Name derived from {{"C", 1}, {"C", 4}, {"C", 4}} as gc = {"C", 9}
However it is worth noting that if requirements contained {"C", 6} and {"C", 3} instead of {"C", 9}, this particular set of collections would fail to satisfy the predicates.
Now to the question finally.
What is the best way to form this into a linq expression prioritizing speed (least iterations)?
The unsolved subset has been re-asked here
Here's my sketch for a linqy solution, but it doesn't address #3 at all. It works by grouping and joining on names. The hard part would then be to determine if there is some matching of requirements to candidates that satisfies the group.
void Main() {
var requirements = new [] {
new NameCount{ Name = "A", Count = 0 },
new NameCount{ Name = "B", Count = 0 },
new NameCount{ Name = "C", Count = 9 },
new NameCount{ Name = "D", Count = 3 },
new NameCount{ Name = "D", Count = 5 },
};
var candidates = new[] {
new NameCount {Name = "B", Count = 0},
new NameCount {Name = "C", Count = 1},
new NameCount {Name = "A", Count = 0},
new NameCount {Name = "D", Count = 2},
new NameCount {Name = "C", Count = 4},
new NameCount {Name = "C", Count = 4}
};
var matched = requirements
.GroupBy(r => r.Name)
.GroupJoin(candidates, rg => rg.Key, c => c.Name,
(rg, cg) => new { requirements = rg, candidates = cg });
bool satisfied = matched.All( /* ??? */ );
}
struct NameCount {
public string Name;
public int Count;
}
For the given input, matched would be this:
.GroupJoin has much better performance characteristics than candidates.Where in the projection.
After reconsidering the revised requirements, I've come up with a invariant assertions that must hold for a solution to exist..
For each paired cg and rg...
|cg.Name| >= |rg.Name|
cg.SummedCount >= rg.SummedCount
Assuming we have satisfied those conditions, a solution MAY exist.
My intuition suggests something similar to the following algorithm:
For each Name...
Let us call each r in rg a basket, and each c in cg an apple.
Sort apples in descending order.
We will keep track of which elements we've assigned to each basket in rg (e.g. r₁ is paired with cg₁.) Maintain sortedness in our buckets by ascending order of rₓ.Count - cgₓ.Count. (This value may be negative.)
Now, iterate through our list of apples, starting with the largest, and assign it to the least filled bucket by iterating through rg. If we overfill the first bucket, we continue descending through the list until we encounter a bucket that would remain unfilled if we put that apple in it. We then choose the previous bucket.
That is, we want to minimize the number of apples necessary to fill each bucket, so we prefer a perfect fit to overfilling, and overfilling to underfilling.
This algorithm does not work on the following case:
rg = (6, 5), cg = (3, 2, 2, 2, 2)
The above algorithm produces
r6 = (3, 2, 2), r5 = (2, 2)
whereas the solution ought to be
r6 = (2, 2, 2), r5 = (3, 2)
Going to post the obvious answer here, but I'm looking for something more elegant.
Given candidates as IEnumerable<'a>, project IEnumerable<'a> groupedCandidates from candidates by calling candidates.Where(c=>c.Count != 0).GroupBy(...) by performing a Sum on all elements with the same name.
Then project simpleCandidates from candidates.Except(groupedCandidates, (c,gc)=>c.Name == gc.Name)
Past here it gets fuzzy because candidates may only satisfy a requirement once.
EDIT: This solution does not meet the revised requirements.
I'm not familiar with LINQ, but it looks like you can do this problem in O(n) unless I misunderstand something. There are three steps to completing this problem.
First, construct a list or hashtable counter and populate it by iterating through c. If we use a hashtable, the size of the hashtable will be the length of c so we don't have to resize our hashtable.
for candidate in c:
counter[candidate.name] += candidate.count
We do this in one pass. O(m) where m is the length of c.
With counter constructed, we construct a hashtable by iterating through r.
for requirement in r:
if not h[requirement.name] or not requirement.count >= h[requirement.name]:
h[requirement.name] = requirement.count
Then, we simply iterate through counter and compare counts.
for sum in counter:
assert h[sum.name] and h[sum.name] >= sum.count
We do this in one pass: O(p) where p is the length of counter.
If this algorithm terminates successfully, our constraints are satisfied, and we've completed it in O(m) + O(o) + O(p)
I finally came up with a workable solution
IEnumerable<Glyph> requirements = t.Objectives.Cast<Glyph>().OrderBy(k => k.Name);
IEnumerable<Glyph> candidates = Resources.Cast<Glyph>().OrderBy(k => k.Name);
IEnumerable<Glyph> zeroCountCandidates = candidates.Where(c => c.Count == 0);
IEnumerable<Glyph> zeroCountRequirements = requirements.Where(r => r.Count == 0);
List<Glyph> remainingCandidates = zeroCountCandidates.ToList();
if (zeroCountCandidates.Count() < zeroCountRequirements.Count())
{
return false;
}
foreach (var r in zeroCountRequirements)
{
if (!remainingCandidates.Contains(r))
{
return false;
}
else
{
remainingCandidates.Remove(r);
}
}
IEnumerable<Glyph> nonZeroCountCandidates = candidates.Where(c => c.Count > 0);
IEnumerable<Glyph> nonZeroCountRequirements = requirements.Where(r => r.Count > 0);
var perms = nonZeroCountCandidates.Permutations();
foreach (var perm in perms)
{
bool isViableSolution = true;
remainingCandidates = perm.ToList();
foreach (var requirement in nonZeroCountRequirements)
{
int countThreshold = requirement.Count;
while (countThreshold > 0)
{
if (remainingCandidates.Count() == 0)
{
isViableSolution = false;
break;
}
var c = remainingCandidates[0];
countThreshold -= c.Count;
remainingCandidates.Remove(c);
}
}
if (isViableSolution)
{
return true;
}
}
return false;
Disgusting isn't it?
algorithm:
if any requirement Name doesn't exist in the candidates, return false
for any requirement having Count = 0
if there aren't at least as many candidates
with the same Name and Count, return false
eliminate all exact matches between candidates and requirements
eliminate requirements (and candidates) where the requirement
and all higher requirements have a higher candidate available
for remaining non-zero requirements
find the subset of candidates
that matches the most requirements
and eliminate the requirements (and candidates)
if there are any remaining non-zero requirements
return false
return true because no unmatched requirements remain
sample implementation:
public static bool IsValid(IEnumerable<string> requirementNames,
IList<int> requirementCounts,
IEnumerable<string> candidateNames,
IList<int> candidateCounts)
{
var requirements = requirementNames
.Select((x, i) => new
{
Name = x,
Count = requirementCounts[i]
})
.ToList();
var candidates = candidateNames
.Select((x, i) => new
{
Name = x,
Count = candidateCounts[i]
})
.ToList();
var zeroRequirements = requirements
.Where(x => x.Count == 0)
.Select(x => x.Name)
.GroupBy(x => x)
.ToDictionary(x => x.Key, x => x.Count());
var zeroCandidates = candidates
.Where(x => x.Count == 0)
.Select(x => x.Name)
.GroupBy(x => x)
.ToDictionary(x => x.Key, x => x.Count());
if (zeroRequirements.Keys.Any(x => !zeroCandidates.ContainsKey(x) ||
zeroCandidates[x] < zeroRequirements[x]))
{
return false;
}
var nonZeroRequirements = requirements
.Where(x => x.Count != 0)
.GroupBy(x => x.Name)
.ToDictionary(x => x.Key,
x => x.Select(y => y.Count)
.GroupBy(y => y)
.ToDictionary(y => y.Key, y => y.Count()));
var nonZeroCandidates = candidates
.Where(x => x.Count != 0)
.GroupBy(x => x.Name)
.ToDictionary(x => x.Key,
x => x.Select(y => y.Count)
.GroupBy(y => y)
.ToDictionary(y => y.Key, y => y.Count()));
foreach (var name in nonZeroRequirements.Keys.ToList())
{
var requirementsForName = nonZeroRequirements[name];
Dictionary<int, int> candidatesForName;
if (!nonZeroCandidates.TryGetValue(name, out candidatesForName))
{
return false;
}
if (candidatesForName.Sum(x => x.Value) <
requirementsForName.Sum(x => x.Value))
{
return false;
}
if (candidatesForName.Sum(x => x.Value*x.Key) <
requirementsForName.Sum(x => x.Value*x.Key))
{
return false;
}
EliminateExactMatches(candidatesForName, requirementsForName);
EliminateHighRequirementsWithAvailableHigherCandidate(candidatesForName, requirementsForName);
EliminateRequirementsThatHaveAMatchingCandidateSum(candidatesForName, requirementsForName);
if (requirementsForName
.Any(x => x.Value > 0))
{
return false;
}
}
return true;
}
private static void EliminateRequirementsThatHaveAMatchingCandidateSum(
IDictionary<int, int> candidatesForName,
IDictionary<int, int> requirementsForName)
{
var requirements = requirementsForName
.Where(x => x.Value > 0)
.OrderByDescending(x => x.Key)
.SelectMany(x => Enumerable.Repeat(x.Key, x.Value))
.ToList();
if (!requirements.Any())
{
return;
}
// requirements -> candidates used
var items = GenerateCandidateSetsThatSumToOrOverflow(
requirements.First(),
candidatesForName,
new List<int>())
.Concat(new[] {new KeyValuePair<int, IList<int>>(0, new List<int>())})
.Select(x => new KeyValuePair<IList<int>, IList<int>>(
new[] {x.Key}, x.Value));
foreach (var count in requirements.Skip(1))
{
var count1 = count;
items = (from i in items
from o in GenerateCandidateSetsThatSumToOrOverflow(
count1,
candidatesForName,
i.Value)
select
new KeyValuePair<IList<int>, IList<int>>(
i.Key.Concat(new[] {o.Key}).OrderBy(x => x).ToList(),
i.Value.Concat(o.Value).OrderBy(x => x).ToList()))
.GroupBy(
x => String.Join(",", x.Key.Select(y => y.ToString()).ToArray()) + ">"
+ String.Join(",", x.Value.Select(y => y.ToString()).ToArray()))
.Select(x => x.First());
}
var bestSet = items
.OrderByDescending(x => x.Key.Count(y => y > 0)) // match the most requirements
.ThenByDescending(x => x.Value.Count) // use the most candidates
.ToList();
var best = bestSet.First();
foreach (var requirementCount in best.Key.Where(x => x > 0))
{
requirementsForName[requirementCount] -= 1;
}
foreach (var candidateCount in best.Value.Where(x => x > 0))
{
candidatesForName[candidateCount] -= 1;
}
}
private static void EliminateHighRequirementsWithAvailableHigherCandidate(
IDictionary<int, int> candidatesForName,
IDictionary<int, int> requirementsForName)
{
foreach (var count in requirementsForName
.Where(x => x.Value > 0)
.OrderByDescending(x => x.Key)
.Select(x => x.Key)
.ToList())
{
while (requirementsForName[count] > 0)
{
var count1 = count;
var largerCandidates = candidatesForName
.Where(x => x.Key > count1)
.OrderByDescending(x => x.Key)
.ToList();
if (!largerCandidates.Any())
{
return;
}
var largerCount = largerCandidates.First().Key;
requirementsForName[count] -= 1;
candidatesForName[largerCount] -= 1;
}
}
}
private static void EliminateExactMatches(
IDictionary<int, int> candidatesForName,
IDictionary<int, int> requirementsForName)
{
foreach (var count in requirementsForName.Keys.ToList())
{
int numberOfCount;
if (candidatesForName.TryGetValue(count, out numberOfCount) &&
numberOfCount > 0)
{
var toRemove = Math.Min(numberOfCount, requirementsForName[count]);
requirementsForName[count] -= toRemove;
candidatesForName[count] -= toRemove;
}
}
}
private static IEnumerable<KeyValuePair<int, IList<int>>> GenerateCandidateSetsThatSumToOrOverflow(
int sumNeeded,
IEnumerable<KeyValuePair<int, int>> candidates,
IEnumerable<int> usedCandidates)
{
var usedCandidateLookup = usedCandidates
.GroupBy(x => x)
.ToDictionary(x => x.Key, x => x.Count());
var countToIndex = candidates
.Select(x => Enumerable.Range(
0,
usedCandidateLookup.ContainsKey(x.Key)
? x.Value - usedCandidateLookup[x.Key]
: x.Value)
.Select(i => new KeyValuePair<int, int>(x.Key, i)))
.SelectMany(x => x)
.ToList();
// sum to List of <count,index>
var sumToElements = countToIndex
.Select(a => new KeyValuePair<int, IList<KeyValuePair<int, int>>>(
a.Key, new[] {a}))
.ToList();
countToIndex = countToIndex.Where(x => x.Key < sumNeeded).ToList();
while (sumToElements.Any())
{
foreach (var set in sumToElements
.Where(x => x.Key >= sumNeeded))
{
yield return new KeyValuePair<int, IList<int>>(
sumNeeded,
set.Value.Select(x => x.Key).ToList());
}
sumToElements = (from a in sumToElements.Where(x => x.Key < sumNeeded)
from b in countToIndex
where !a.Value.Any(x => x.Key == b.Key && x.Value == b.Value)
select new KeyValuePair<int, IList<KeyValuePair<int, int>>>(
a.Key + b.Key,
a.Value.Concat(new[] {b})
.OrderBy(x => x.Key)
.ThenBy(x => x.Value)
.ToList()))
.GroupBy(x => String.Join(",", x.Value.Select(y => y.Key.ToString()).ToArray()))
.Select(x => x.First())
.ToList();
}
}
private static IEnumerable<int> GetAddendsFor(int sum, Random random)
{
var values = new List<int>();
while (sum > 0)
{
var addend = random.Next(1, sum);
sum -= addend;
values.Add(addend);
}
return values;
}
Tests:
[Test]
public void ABCC_0063__with_candidates__BCADCC_010244__should_return_false()
{
var requirementNames = "ABCC".Select(x => x.ToString()).ToArray();
var requirementCounts = new[] {0, 0, 6, 3};
var candidateNames = "BCADCC".Select(x => x.ToString()).ToArray();
var candidateCounts = new[] {0, 1, 0, 2, 4, 4};
var actual = IsValid(requirementNames, requirementCounts, candidateNames, candidateCounts);
actual.ShouldBeFalse();
}
[Test]
public void ABC_003__with_candidates__BCADCC_010244__should_return_true()
{
var requirementNames = "ABC".Select(x => x.ToString()).ToArray();
var requirementCounts = new[] {0, 0, 3};
var candidateNames = "BCADCC".Select(x => x.ToString()).ToArray();
var candidateCounts = new[] {0, 1, 0, 2, 4, 4};
var actual = IsValid(requirementNames, requirementCounts, candidateNames, candidateCounts);
actual.ShouldBeTrue();
}
[Test]
public void ABC_003__with_candidates__BCAD_0102__should_return_false()
{
var requirementNames = "ABC".Select(x => x.ToString()).ToArray();
var requirementCounts = new[] {0, 0, 3};
var candidateNames = "BCAD".Select(x => x.ToString()).ToArray();
var candidateCounts = new[] {0, 1, 0, 2};
var actual = IsValid(requirementNames, requirementCounts, candidateNames, candidateCounts);
actual.ShouldBeFalse();
}
[Test]
public void ABC_009__with_candidates__BCADCC_010244__should_return_true()
{
var requirementNames = "ABC".Select(x => x.ToString()).ToArray();
var requirementCounts = new[] {0, 0, 9};
var candidateNames = "BCADCC".Select(x => x.ToString()).ToArray();
var candidateCounts = new[] {0, 1, 0, 2, 4, 4};
var actual = IsValid(requirementNames, requirementCounts, candidateNames, candidateCounts);
actual.ShouldBeTrue();
}
[Test]
public void FuzzTestIt()
{
var random = new Random();
const string names = "ABCDE";
for (var tries = 0; tries < 10000000; tries++)
{
var numberOfRequirements = random.Next(5);
var shouldPass = true;
var requirementNames = new List<string>();
var requirementCounts = new List<int>();
var candidateNames = new List<string>();
var candidateCounts = new List<int>();
for (var i = 0; i < numberOfRequirements; i++)
{
var name = names.Substring(random.Next(names.Length), 1);
switch (random.Next(6))
{
case 0: // zero-requirement with corresponding candidate
requirementNames.Add(name);
requirementCounts.Add(0);
candidateNames.Add(name);
candidateCounts.Add(0);
break;
case 1: // zero-requirement without corresponding candidate
requirementNames.Add(name);
requirementCounts.Add(0);
shouldPass = false;
break;
case 2: // non-zero-requirement with corresponding candidate
{
var count = random.Next(1, 10);
requirementNames.Add(name);
requirementCounts.Add(count);
candidateNames.Add(name);
candidateCounts.Add(count);
}
break;
case 3: // non-zero-requirement with matching sum of candidates
{
var count = random.Next(1, 10);
requirementNames.Add(name);
requirementCounts.Add(count);
foreach (var value in GetAddendsFor(count, random))
{
candidateNames.Add(name);
candidateCounts.Add(value);
}
}
break;
case 4: // non-zero-requirement with matching overflow candidate
{
var count = random.Next(1, 10);
requirementNames.Add(name);
requirementCounts.Add(count);
candidateNames.Add(name);
candidateCounts.Add(count + 2);
}
break;
case 5: // non-zero-requirement without matching candidate or sum or candidates
{
var count = random.Next(10, 20);
requirementNames.Add(name);
requirementCounts.Add(count);
shouldPass = false;
}
break;
}
}
try
{
var actual = IsValid(requirementNames, requirementCounts, candidateNames, candidateCounts);
actual.ShouldBeEqualTo(shouldPass);
}
catch (Exception e)
{
Console.WriteLine("Requirements: " + String.Join(", ", requirementNames.ToArray()));
Console.WriteLine(" " +
String.Join(", ", requirementCounts.Select(x => x.ToString()).ToArray()));
Console.WriteLine("Candidates: " + String.Join(", ", candidateNames.ToArray()));
Console.WriteLine(" " +
String.Join(", ", candidateCounts.Select(x => x.ToString()).ToArray()));
Console.WriteLine(e);
Assert.Fail();
}
}
}
var ids = new int[] { 3, 2, 20, 1 };
var entities = categories.Where(entity => ids.Contains(entity.Id));
I have to sort entities by exactly same like in ids array. How can i do that ?
This should do the trick (written off the top of my head, so may have mistakes)
var ids = new int[] { 3, 2, 20, 1 };
var ordering = ids.Select((id,index) => new {id,index});
var entities =
categories
.Where(entity => ids.Contains(entity.Id))
.AsEnumerable() //line not necessary if 'categories' is a local sequence
.Join(ordering, ent => ent.Id, ord => ord.id, (ent,ord) => new {ent,ord})
.OrderBy(x => x.ord.index)
.Select(x => x.ent)
You could use OrderBy with the index of the Ids in ids.
To get the index of an Id from ids, you could create a map of Id to index. That way you can look up the index in almost constant time, instead of having to call IndexOf and traverse the whole list each time.
Something like this:
var idToIndexMap = ids
.Select((i, v) => new { Index = i, Value = v })
.ToDictionary(
pair => pair.i,
pair => pair.v
);
var sortedEntities = categories
.Where(e => ids.Contains(e.Id))
.ToList() // Isn't necessary if this is Linq-to-Objects instead of entities...
.OrderBy(e => idToIndexMap[e.Id])
;
You may have a go with this:
public class Foo
{
public void Bar()
{
int[] idOrder = new int[] { 3, 2, 20, 1 };
var lookup = idOrder.ToDictionary(i => i,
i => Array.IndexOf(idOrder, i));
foreach(var a in idOrder.OrderBy(i => new ByArrayComparable<int>(lookup, i)))
Console.WriteLine(a);
}
}
public class ByArrayComparable<T> : IComparable<ByArrayComparable<T>> where T : IComparable<T>
{
public readonly IDictionary<T, int> order;
public readonly T element;
public ByArrayComparable(IDictionary<T, int> order, T element)
{
this.order = order;
this.element = element;
}
public int CompareTo(ByArrayComparable<T> other)
{
return this.order[this.element].CompareTo(this.order[other.element]);
}
}
This works for unique elements only, but the lookup efford is constant.
I am trying to use LINQ to return the an element which occurs maximum number of times AND the number of times it occurs.
For example:
I have an array of strings:
string[] words = { "cherry", "apple", "blueberry", "cherry", "cherry", "blueberry" };
//...
Some LINQ statement here
//...
In this array, the query would return cherry as the maximum occurred element, and 3 as the number of times it occurred. I would also be willing to split them into two queries if that is necessary (i.e., first query to get the cherry, and second to return the count of 3.
The solutions presented so far are O(n log n). Here's an O(n) solution:
var max = words.GroupBy(w => w)
.Select(g => new { Word = g.Key, Count = g.Count() })
.MaxBy(g => g.Count);
Console.WriteLine(
"The most frequent word is {0}, and its frequency is {1}.",
max.Word,
max.Count
);
This needs a definition of MaxBy. Here is one:
public static TSource MaxBy<TSource>(
this IEnumerable<TSource> source,
Func<TSource, IComparable> projectionToComparable
) {
using (var e = source.GetEnumerator()) {
if (!e.MoveNext()) {
throw new InvalidOperationException("Sequence is empty.");
}
TSource max = e.Current;
IComparable maxProjection = projectionToComparable(e.Current);
while (e.MoveNext()) {
IComparable currentProjection = projectionToComparable(e.Current);
if (currentProjection.CompareTo(maxProjection) > 0) {
max = e.Current;
maxProjection = currentProjection;
}
}
return max;
}
}
var topWordGroup = words.GroupBy(word => word).OrderByDescending(group => group.Count()).FirstOrDefault();
// topWordGroup might be a null!
string topWord = topWordGroup.Key;
int topWordCount = topWordGroup.Count;
And in case if we don't like O(N log N):
var topWordGroup = words.GroupBy(word => word).Aggregate((current, acc) => current.Count() < acc.Count() ? acc : current);
First thing that comes to mind (meaning there is probably a more efficient way)
var item = words.GroupBy(x => x).OrderByDescending(x => x.Count()).First()
//item.Key is "cherry", item.Count() is 3
EDIT: forgot op wanted the name and the count
string[] words = { "cherry", "apple", "blueberry", "cherry", "cherry", "blueberry" };
var topWordAndCount=words
.GroupBy(w=>w)
.OrderByDescending(g=>g.Count())
.Select(g=>new {Word=g.Key,Count=g.Count()})
.FirstOrDefault();
//if(topWordAndCount!=null)
//{
// topWordAndCount.Word
// topWordAndCount.Count
Try this one:
Converting SQL containing top, count, group and order to LINQ (2 Entities)
string[] words = { "cherry", "apple", "blueberry", "cherry", "cherry", "blueberry" };
var r = words.GroupBy (x => x)
.OrderByDescending (g => g.Count ())
.FirstOrDefault ();
Console.WriteLine (String.Format ("The element {0} occurs {1} times.", r.Key, r.Count ()));
A simpler O(n) solution:
var groups = words.GroupBy(x => x);
var max = groups.Max(x => x.Count());
var top = groups.First(y => y.Count() == max).Key;
Here's a very fast O(n) solution in one line(!):
s.GroupBy(x => x).Aggregate((IGrouping<string,string>)null, (x, y) => (x != null && y != null && x.Count() >= y.Count()) || y == null ? x : y, x => x);
Or this:
s.GroupBy(x => x).Select(x => new { Key = x.Key, Count = x.Count() }).Aggregate(new { Key = "", Count = 0 }, (x, y) => x.Count >= y.Count ? x : y, x => x);
List<int> a = 11,2,3,11,3,22,9,2
//output
11
This may not be the most efficient way, but it will get the job done.
public static int MostFrequent(IEnumerable<int> enumerable)
{
var query = from it in enumerable
group it by it into g
select new {Key = g.Key, Count = g.Count()} ;
return query.OrderByDescending(x => x.Count).First().Key;
}
And the fun single line version ...
public static int MostFrequent(IEnumerable<int> enumerable)
{
return (from it in enumerable
group it by it into g
select new {Key = g.Key, Count = g.Count()}).OrderByDescending(x => x.Count).First().Key;
}
a.GroupBy(item => item).
Select(group => new { Key = group.Key, Count = group.Count() }).
OrderByDescending(pair => pair.Count).
First().
Key;
Another example :
IEnumerable<int> numbers = new[] { 11, 2, 3, 11, 3, 22, 9, 2 };
int most = numbers
.Select(x => new { Number = x, Count = numbers.Count(y => y == x) })
.OrderByDescending(z => z.Count)
.First().Number;