I have a list of values like this
1000, 20400
22200, 24444
The ranges don\'t overlap.
What I want to do is have a c# function that can
Assuming your ranges don't overlap:
-> Put all your range numbers in an array.
-> Sort your array.
-> Also keep a HashSet for your startvalues.
-> Now do a binary search on your number. Two possiblities:
--> Array range left of (smaller then) your number is a start value: your number is in range.
--> Array range left of (smaller then) your number is not a start value: your number is not in range.
I'd try the simplest option first, and optimize if that doesn't meet your needs.
class Range {
int Lower { get; set; }
int Upper { get; set; }
}
List<Range>.FirstOrDefault(r => i >= r.Lower && i <= r.Upper);
Is this functionally what you're after? If so, and you just want it to be more performant, than change the foreach in the ValueRangeCollection to a binary search..
public struct ValueRange
{
public int LowVal;
public int HiVal;
public bool Contains (int CandidateValue)
{ return CandidateValue >= LowVal && CandidateValue <= HiVal; }
public ValueRange(int loVal, int hiVal)
{
LowVal = loVal;
HiVal = hiVal;
}
}
public class ValueRangeCollection: SortedList<int, ValueRange>
{
public bool Contains(int candValue)
{
foreach ( ValueRange valRng in Values)
if (valRng.Contains(candValue)) return true;
return false;
}
public void Add(int loValue, int hiValue)
{
Add(loValue, new ValueRange(loValue, hiValue));
}
}
As previously mentioned, if the set of ranges are big and non-overlapping, it is best to do a binary search. One way to do this is to use SortedDictionary
, which implements a red-black tree to give a O(log(n)) search time. We can use the ranges as keys, and do a dictionary lookup by converting the single value we want to match into a range of a single point. If we implement the CompareTo
method so that ranges that are overlapping are considered equal/matching, the dictionary lookup will find the matching range for use.
public struct Range : IComparable<Range>
{
public int From;
public int To;
public Range(int point)
{
From = point;
To = point;
}
public Range(int from, int to)
{
From = from;
To = to;
}
public int CompareTo(Range other)
{
// If the ranges are overlapping, they are considered equal/matching
if (From <= other.To && To >= other.From)
{
return 0;
}
// Since the ranges are not overlapping, we can compare either end
return From.CompareTo(other.From);
}
}
public class RangeDictionary
{
private static SortedDictionary<Range, string> _ranges = new SortedDictionary<Range, string>();
public RangeDictionary()
{
_ranges.Add(new Range(1, 1000), "Alice");
_ranges.Add(new Range(1001, 2000), "Bob");
_ranges.Add(new Range(2001, 3000), "Carol");
}
public string Lookup(int key)
{
/* We convert the value we want to lookup into a range,
* so it can be compared with the other ranges */
var keyAsRange = new Range(key);
string value;
if (_ranges.TryGetValue(keyAsRange, out value))
{
return value;
}
return null;
}
}
As an example, running the following code
var ranges = new RangeDictionary();
var value = ranges.Lookup(1356);
value
will in this case contain the string "Bob"
, since 1356 matches the range 1001-2000.
In your case, if you are interested in fetching the range itself, you can use the range as both key and value in the dictionary. The example code can be easily extended to holding generic values instead.
As a sidenote, this trick can also be done using a SortedList
with virtually the same code, which uses less memory (array instead of tree) but have slower insertion/deletion time for unsorted data. They both use the default comparator for the key type (or a specified one) to compare values. The normal C# Dictionary
on the other hand uses GetHashCode
and Equals
to compare values.
class Range
{
public int Start { get; set; }
public int End { get; set; }
static Dictionary<int, Range> values;
static int[] arrToBinarySearchIn;
public static void BuildRanges(IEnumerable<Range> ranges) {
values = new Dictionary<int, Range>();
foreach (var item in ranges)
values[item.Start] = item;
arrToBinarySearchIn = values.Keys.ToArray();
Array.Sort(arrToBinarySearchIn);
}
public static Range GetRange(int value)
{
int searchIndex = Array.BinarySearch(arrToBinarySearchIn, value);
if (searchIndex < 0)
searchIndex = ~searchIndex - 1;
if (searchIndex < 0)
return null;
Range proposedRange = values[arrToBinarySearchIn[searchIndex]];
if (proposedRange.End >= value)
return proposedRange;
return null;
}
}
class Ranges
{
int[] starts = new[] { 1000, 22200 };
int[] ends = new[] { 20400, 24444 };
public int RangeIndex(int test)
{
int index = -1;
if (test >= starts[0] && test <= ends[ends.Length - 1])
{
index = Array.BinarySearch(ends, test);
if (index <= 0)
{
index = ~index;
if (starts[index] > test) index = -1;
}
}
return index;
}
}
Obviously, how you instantiate the class is up to you. Maybe pass in a DataTable and construct the arrays from that.