We Keep Coding

Creating dynamic range aggregations in Nest

I have a set of POCO facets with ranges, that I have received from a client. I need to convert these ultimately into an AggregationDictionary. I cannot figure out the syntax for creating a dynamic set of aggregations (possilbly of type RangeAggregationDescriptor) and need help with this.
My POCO objects are below:
public class TypedFacets
{
public string Name { get; set; }
public string Field { get; set; }
public IReadOnlyCollection<Range> RangeValues { get; set; } = new List<Range>();
public int Size { get; set; }
}
public class Range
{
public string Name { get; set; }
public double? From { get; set; }
public double? To { get; set; }
}
The Nest generation looks like below:
var facets = new List<TypedFacets>()
{
new TypedFacets()
{
Name = "potatoRange",
Field = "potatoRange",
RangeValues = new List<Range>()
{
new Range()
{
From = 0,
To = null,
Name = "chips"
},
new Range()
{
From = 1,
To = null,
Name = "crisps"
}
}
}
};
var aggregations = new AggregationContainerDescriptor<Template>();
facets.Where(f => f.RangeValues.Any()).ToList().ForEach(f =>
{
var rad = new RangeAggregationDescriptor<Template>();
f.RangeValues.ToList().ForEach(rangeValue =>
{
rad = rad.Ranges(rs => rs.From(rangeValue.From).To(rangeValue.To).Key(rangeValue.Name));
});
// this line doesn't work and needs to change
aggregations.Range(f.Name, r => r
.Field(f.Field).Ranges(rs => rad.Ranges));
});
return ((IAggregationContainer)aggregations).Aggregations;
I'm not sure how to fix the above. Any help would be appreciated.

I eventually found the solution for this. You can create the dynamic ranges as per below
private Func<AggregationRangeDescriptor, IAggregationRange>[] CreateRangeRanges(TypedFacets rangedAgg)
{
var rangeRanges = new List<Func<AggregationRangeDescriptor, IAggregationRange>>();
rangedAgg.RangeValues.ToList().ForEach(rangeValue =>
{
rangeRanges.Add(rs => rs.From(rangeValue.From).To(rangeValue.To).Key(rangeValue.Name));
});
return rangeRanges.ToArray();
}
And then assing them like below
facets.Where(f => f.RangeValues.Any()).ToList().ForEach(f =>
{
aggregations.Range(f.Name, r => r
.Field(f.Field).Ranges(CreateRangeRanges(f)));
});

Updating child items in List updates all Items in C#

Here are my list :
public class PayRateDaysModel
{
public string day_name { get; set; }
public List<RateList> multiplier { get; set; }
}
public class RateList
{
public double start_after { get; set; }
public double rate_multiplier { get; set; }
}
when I'm trying to update any value in multiplier of PayRateDaysModel then its updating all multiplier values of PayRateDaysModel. Iw ant to update only current item. Below is my code :
var dayExists = daysModel.Where(x => x.day_name == day_name).FirstOrDefault();
if(dayExists==null)
{
PayRateDaysModel days = new PayRateDaysModel();
days.day_name = day_name;
days.multiplier = rate_list;
daysModel.Add(days);
}
else
{
//update
dayExists.day_name = "abc";
dayExists.multiplier.FirstOrDefault().rate_multiplier = 1;
}

Based on what you have shown to us I would think that you are creating the rate_list somewhere above like rate_list = new RateList(…) and you are setting this to all of your days in days.multiplier = rate_list;. Since you did not recreate that rate_list for every element, any time you change it in one of your dayExists you will change it for all the others as well.
So you should do something like this days.multiplier = new RateList(…);

Increment through property names in object

I have the following object:
namespace BluetoothExample
{
public class Assay
{
public double Band_1 //Vis450
{
get;
set;
}
public double Band_2 //Vis500
{
get;
set;
}
public double Band_3 //Vis550
{
get;
set;
}
public double Band_4 //Vis570
{
get;
set;
}
}
}
I want to populate the 4 bands in my object. Which I currently do in the following way:
public Populate()
{
int _i = 0;
double[] nirData = new double[4];
MyDevice.Characteristic.ValueUpdated += (sender, e) =>
{
nirData[_i] = BitConverter.ToDouble(e.Characteristic.Value, 0);
_i++;
};
Assay assay = new Assay();
assay.Band_1 = nirData[0];
assay.Band_2 = nirData[1];
assay.Band_3 = nirData[2];
assay.Band_4 = nirData[3];
}
I was wondering if it was possible to do the entire thing inside the MyDevice.Characteristic.ValueUpdate method instead? My thought is that it should be possible to increment and populate the properties of my object like so:
string name = "assay.Band_" + _i;
name = BitConverter.ToDouble(e.Characteristic.Value, 0);
This is obviously wrong, but it sort of demonstrates my idea.

Just make your band an Array, or List:
public class Assay
{
public double[] Band
{
get;
set;
}
}
And then you can simply assign it like:
public Populate()
{
int _i = 0;
double[] nirData = new double[4];
MyDevice.Characteristic.ValueUpdated += (sender, e) =>
{
nirData[_i] = BitConverter.ToDouble(e.Characteristic.Value, 0);
_i++;
};
Assay assay = new Assay();
assay.Band = nirData;
}

Object creation & equivalent attribute signatures. Does this satisfy the requirements?

I'm trying to upgrade my general class design skills in C#, and want you guys to reveal code-smells I might have. (hope general discussions are allowed at Stackoverflow.com)
Regarding to This Tutorial, I wrapped this up to C# and tried to chain constructors, avoid same attribute signatures and solve it the right way, without duplicating code.
The problem in this tutorial: The Running Back and Wide Receiver would use the same attribute signatures. Multiple constructions for each player role would cause errors. Instead you might use object creation.
Below, how I solved it.
Football Player class:
public class FootballPlayer
{
public double PasserRating { get; set; } // Specific to QBs
public int RushingYards { get; set; } // Specific to RBs & QBs
public int ReceivingYards { get; set; } // Specific to RBs & WRs
public int TotalTackles { get; set; } // Specific to DEF
public int Interceptions { get; set; } // Specific to DEF
public int FieldGoals { get; set; } // Specific to Kickers
public double AvgPunt { get; set; } // Specific to Punters
public double AvgKickoffReturn { get; set; }// Specific to Special Teams
public double AvgPuntreturn { get; set; } // Specific to Special Teams
private FootballPlayer() : this(0.0, 0, 0, 0, 0, 0, 0.0, 0.0, 0.0) { }
private FootballPlayer(double passerRating, int rushingYards, int receivingYards, int totalTackles, int interceptions, int fieldGoals, double avgPunt, double avgKickoffReturn, double avgPuntreturn)
{
PasserRating = passerRating;
RushingYards = rushingYards;
ReceivingYards = receivingYards;
TotalTackles = totalTackles;
Interceptions = interceptions;
FieldGoals = fieldGoals;
AvgPunt = avgPunt;
AvgKickoffReturn = avgKickoffReturn;
AvgPuntreturn = avgPuntreturn;
}
public static FootballPlayer CreateQb(double passerRating, int rushingYards)
{
return new FootballPlayer
{
PasserRating = passerRating,
RushingYards = rushingYards
};
}
public static FootballPlayer CreateRb(int rushingYards)
{
return new FootballPlayer
{
RushingYards = rushingYards
};
}
public static FootballPlayer CreateWr(int receivingYards)
{
return new FootballPlayer
{
ReceivingYards = receivingYards
};
}
}
Program:
class Program
{
static void Main()
{
var aaronRodgers = FootballPlayer.CreateQb(132.0, 259);
var drewBrees = FootballPlayer.CreateQb(129.6, 244);
Console.WriteLine("Aaron Rodgers passer Rating: {0}", aaronRodgers.PasserRating);
// OUTPUT: Aaron Rodgers passer Rating: 132
Console.WriteLine("Drew Brees passer Rating: {0}", drewBrees.PasserRating);
// OUTPUT: Drew Brees passer Rating: 129.6
Console.ReadLine();
}
}
Would you guys say that's the way to go?
Or what should I take into account?

C# AStar problems, won't do it correctly

I am currently working on A* pathfinding, but I am having some problems.
It does the wrong path before taking the best path to the end.
What am I doing wrong?
Source code: http://basic.apayostudios.com/AStar.zip
Online:
Game.cs http://pastie.org/1656955
Node.cs http://pastie.org/1656956
Enums:
public enum NodeType
{
None,
Solid,
Start,
End
}
Thanks!

Here's my A-star path finder, it works and you're free to use it for learning and comparing your solution against it, rip me off if you like. But there are dependencies that are missing from this code and I'm using fixed-point arithmetic. This won't build without some changes. This code is relatively high level and should be easy enough to reverse engineer.
public class AgentPathfinder
{
class SolutionComparer : IComparer<Solution>
{
public int Compare(Solution x, Solution y)
{
return x.Cost.CompareTo(y.Cost);
}
}
class Solution
{
public List<FixedVector> Path { get; private set; }
public FixedVector LastPosition { get { return Path[Path.Count - 1]; } }
public Fixed32 Heuristic { get; set; }
public Fixed32 Cost { get; set; }
public Solution(FixedVector position, Fixed32 heuristic)
{
Path = new List<FixedVector>(2) { position };
Heuristic = heuristic;
Cost = Path.Count + heuristic;
}
public Solution(FixedVector position
, Fixed32 heuristic
, List<FixedVector> path)
{
Path = new List<FixedVector>(path) { position };
Heuristic = heuristic;
Cost = Path.Count + heuristic;
}
}
// TODO: replace with pathable terrain data
public Map Map { get; set; }
public FixedVector Position { get; set; }
public FixedVector Destination { get; set; }
public List<FixedVector> Path { get; set; }
public void Compute()
{
var visited = new bool[(int)Map.Size.Width, (int)Map.Size.Height];
var pq = new PriorityQueue<Solution>(new SolutionComparer());
var bestFit = new Solution(new FixedVector((int)(Position.X + 0.5)
, (int)(Position.Y + 0.5))
, (Destination - Position).Length
);
pq.Enqueue(bestFit);
while (pq.Count > 0)
{
var path = pq.Dequeue(); // optimal, thus far
if (path.Heuristic < bestFit.Heuristic) // best approximation?
{
// if we starve all other paths we
// fallback to this, which should be the best
// approximation for reaching the goal
bestFit = path;
}
for (int i = 0; i < FixedVector.Adjacent8.Length; i++)
{
var u = path.LastPosition + FixedVector.Adjacent8[i];
if (Map.Size.Contains(u))
{
if (Map.IsPathable(u))
{
if (!visited[(int)u.X, (int)u.Y])
{
// heuristic (straight-line distance to the goal)
var h = (Destination - u).Length;
var solution = new Solution(u, h, path.Path);
if (h < 1)
{
Path = solution.Path;
return; // OK, done
}
else
{
// keep looking
pq.Enqueue(solution);
}
visited[(int)u.X, (int)u.Y] = true;
}
}
}
}
}
Path = bestFit.Path;
}
}