public static void Main( string[] args )
{
InternalData internalData = new InternalData();
// output string representation of internalData
Console.WriteLine( "After instantiation:\n{0}", internalData );
// change internal access data in internalData
internalData.number = 77;
internalData.message = "Goodbye";
// output string representation of internalData
Console.WriteLine( "\nAfter changing values:\n{0}", internalData );
}
public void Calculator_Calculate_OneDoctorWithoutPacients()
{
var data = new InternalData();
var doc = new InternalDataDoctor() { DoctorId = 1 };
data.Doctors.Add(doc);
var res0 = Calc.Calculate(data, CalculationType.ByDoctor, excludeDoctorsWithoutPacients: false, excludePacientsWithoutMeasurements: false);
Assert.AreEqual(1, res0.Count(), "res0 must contains 1 result");
Assert.AreEqual(0d, res0.First().Result);
var res1 = Calc.Calculate(data, CalculationType.ByDoctor, excludeDoctorsWithoutPacients: true, excludePacientsWithoutMeasurements: false);
Assert.AreEqual(0, res1.Count(), "res1 must contains no result");
data.Doctors.Add(new InternalDataDoctor() { DoctorId = 2 });
var res2 = Calc.Calculate(data, CalculationType.ByDoctor, excludeDoctorsWithoutPacients: false, excludePacientsWithoutMeasurements: false);
Assert.AreEqual(2, res2.Count(), "res2 must contains 2 result");
var res3 = Calc.Calculate(data, CalculationType.ByDoctor, excludeDoctorsWithoutPacients: true, excludePacientsWithoutMeasurements: false);
Assert.AreEqual(0, res3.Count(), "res3 must contains 0 result");
}
public void Calculator_Calculate_OneDoctorWithOnePacientsAndMeasurements()
{
var data = new InternalData();
var doc = new InternalDataDoctor() { DoctorId = 1 };
var pac0 = new InternalDataPacient() { DoctorId = doc.DoctorId, PacientId = 1 };
data.Doctors.Add(doc);
data.Pacients.Add(pac0);
double start_height = 100;
double height = start_height;
var dt = DateTime.Now;
for (int i=0; i<100; i++)
{
var me = new InternalDataMeasurement() { PacientId = pac0.PacientId, MeasurementId = i };
var hc = new InternalDataHeighComponent() { MeasurementId = me.MeasurementId, Height = height };
var ts = new InternalDataTimestamp() { MeasurementId = me.MeasurementId, Timestamp = dt };
data.Measurements.Add(me);
data.HeighComponent.Add(hc);
data.Timestamps.Add(ts);
height += start_height;
dt = dt.AddDays(7);
}
var res0 = Calc.Calculate(data, CalculationType.ByDoctor, CalculationTimeType.PerWeek, excludeDoctorsWithoutPacients: false, excludePacientsWithoutMeasurements: false);
Assert.AreEqual(start_height, res0.First().Result, string.Format("result for res0 must equals {0} (per week)", start_height));
var res1 = Calc.Calculate(data, CalculationType.ByDoctor, CalculationTimeType.PerDay, excludeDoctorsWithoutPacients: false, excludePacientsWithoutMeasurements: false);
Assert.AreEqual(start_height / 7d, res1.First().Result.Value, 0.01, string.Format("result for res1 must equals {0}/7 (per day)", start_height));
var res2 = Calc.Calculate(data, CalculationType.ByDoctor, CalculationTimeType.PerYear, excludeDoctorsWithoutPacients: false, excludePacientsWithoutMeasurements: false);
Assert.AreEqual(start_height / 7d * 365.25, res2.First().Result.Value, 0.01, string.Format("result for res2 must equals {0}/7*365.25 (per year)", start_height));
var res3 = Calc.Calculate(data, CalculationType.ByDoctor, CalculationTimeType.PerMonth, excludeDoctorsWithoutPacients: false, excludePacientsWithoutMeasurements: false);
Assert.AreEqual(start_height / 7d * 365.25 / 12d, res3.First().Result.Value, 0.01, string.Format("result for res3 must equals {0}/7*365.25/12 (per month)", start_height));
}
/// <summary>
/// 放送情報の取得を開始します。
/// </summary>
private static void BeginGetLiveInfo(long liveId,
InternalData internalData,
AutoResetEvent ev)
{
WebUtil.RequestHttpTextAsync(
NicoString.GetLiveUrl(liveId),
null,
internalData.Cookie,
Encoding.UTF8,
(result, text) =>
{
try
{
// 放送情報が取得できなければエラーとします。
var v = LiveInfo.CreateFromHtml(
string.Format("lv{0}", liveId),
text);
internalData.LiveStreamInfo.LiveInfo = v;
}
catch (Exception ex)
{
internalData.Exception = ex;
}
ev.Set();
});
}
/// <summary>
/// playerstatusの取得を開始します。
/// </summary>
private static void BeginGetPlayerStatus(long liveId,
InternalData internalData,
AutoResetEvent ev)
{
WebUtil.RequestHttpAsync(
NicoString.GetPlayerStatusUrl(liveId),
null,
internalData.Cookie,
(result, data) =>
{
try
{
// ステータスがおかしければエラーとします。
var v = PlayerStatus.CreateFromXml(liveId, GetXml(data));
internalData.LiveStreamInfo.PlayerStatus = v;
}
catch (Exception ex)
{
internalData.Exception = ex;
}
ev.Set();
});
}
private void settingsGrid_PropertyValueChanged(object s, PropertyValueChangedEventArgs e)
{
InternalData.SaveSettings();
}
/* Method: ReadTrainFromFile
* Reads a file that stores training data.
*
* The file must be formatted like:
* >TrainDataLength InputCount OutputCount
* >inputdata seperated by space
* >outputdata seperated by space
* >
* >.
* >.
* >.
* >
* >inputdata seperated by space
* >outputdata seperated by space
*
* See also:
* <NeuralNet::TrainOnData>, <SaveTrain>, <fann_read_train_from_file at http://libfann.github.io/fann/docs/files/fann_train-h.html#fann_read_train_from_file>
*
* This function appears in FANN >= 1.0.0
*/
public bool ReadTrainFromFile(string filename)
{
return(InternalData.read_train_from_file(filename));
}
public void Calculator_Calculate_OneDoctorWithPacients()
{
var data = new InternalData();
var doc = new InternalDataDoctor() { DoctorId = 1 };
var pac0 = new InternalDataPacient() { DoctorId = doc.DoctorId, PacientId = 1 };
var pac1 = new InternalDataPacient() { DoctorId = doc.DoctorId, PacientId = 2 };
data.Doctors.Add(doc);
data.Pacients.Add(pac0);
data.Pacients.Add(pac1);
var res0 = Calc.Calculate(data, CalculationType.ByDoctor, excludeDoctorsWithoutPacients: false, excludePacientsWithoutMeasurements: false);
Assert.AreEqual(1, res0.Count(), "res0 must contains 1 result");
Assert.AreEqual(0d, res0.First().Result, "result for res0 must equals 0");
var res1 = Calc.Calculate(data, CalculationType.ByPacient, excludeDoctorsWithoutPacients: false, excludePacientsWithoutMeasurements: false);
Assert.AreEqual(2, res1.Count(), "res1 must contains 2 result");
Assert.AreEqual(0d, res1.Distinct().First().Result, "result for res1 must equals 0");
var res2 = Calc.Calculate(data, CalculationType.ByPacient, excludeDoctorsWithoutPacients: false, excludePacientsWithoutMeasurements: true);
Assert.AreEqual(0, res2.Count(), "res2 must contains 0 result");
var res3 = Calc.Calculate(data, CalculationType.ByPacient, excludeDoctorsWithoutPacients: true, excludePacientsWithoutMeasurements: true);
Assert.AreEqual(0, res3.Count(), "res3 must contains 2 result");
data.Doctors.Add(new InternalDataDoctor() { DoctorId = 2 });
var res4 = Calc.Calculate(data, CalculationType.ByDoctor, excludeDoctorsWithoutPacients: false, excludePacientsWithoutMeasurements: false);
Assert.AreEqual(2, res4.Count(), "res4 must contains 2 result");
var res5 = Calc.Calculate(data, CalculationType.ByDoctor, excludeDoctorsWithoutPacients: true, excludePacientsWithoutMeasurements: false);
Assert.AreEqual(1, res5.Count(), "res5 must contains 1 result");
}
public void Remove(object key)
{
InternalData.Remove(key);
}
internal SWIGTYPE_p_fann_train_data ToFannTrainData()
{
return(InternalData.to_fann_train_data());
}
private void onReceive(IAsyncResult ar)
{
bool read_complete = false;
int next_read_size = 1024;
try
{
int bytesRead = clientSocket.EndReceive(ar);
if (bytesRead < 1024)
{
byte[] this_read = new byte[bytesRead];
Array.Copy(byteData, 0, this_read, 0, bytesRead);
parts.AddRange(this_read);
if (parts.Count >= 1024)
{
if (parts.Count > 1024)
{
LogEvent($"Warning: Read {parts.Count} bytes, expected 1024");
}
Array.Copy(parts.ToArray(), 0, byteData, 0, 1024);
parts = new List <byte>();
read_complete = true;
}
else
{
//Not done, set the next read size to the remaining
next_read_size = 1024 - parts.Count;
}
}
else if (bytesRead == 1024)
{
//next_read_size = 1024; //set above
//bytedata is fine
read_complete = true;
}
//we have the 1024 to send forward
if (read_complete)
{
byte[] internalData = new byte[0];
if (NetData.ReceiveData("server", byteData, out internalData))
{
InternalData ido = new InternalData(internalData);
ReceiveEvent(ido);
}
}
// Continue to listen
byteData = new byte[1024];
clientSocket.BeginReceive(byteData,
0,
next_read_size,
SocketFlags.None,
new AsyncCallback(onReceive),
null); //should be socket?
}
catch (ObjectDisposedException)
{
// Chances are the object was disposed by Disconnect on another thread
Disconnect("Socket disposed"); //Just in case
}
catch (Exception ex)
{
LogEvent($"Error: Unable to receive message (OnReceive).\r\n{ex}");
Disconnect("Error recieving message. See log for details.");
}
}
public void Dispose()
{
IsDisposed = true;
InternalData.Clear();
}
/// <inheritdoc cref="TableWorker{T}.NotifyGlobalDataChanged"/>
public override void NotifyGlobalDataChanged()
{
var wasDirty = false;
foreach (EventItem item in ToolkitUtils.Data.Events.Select(item => new { item, existing = InternalData.Find(i => i.Data.Equals(item)) })
.Where(t => t.existing == null)
.Select(t => t !.item))
{
InternalData.Add(new TableSettingsItem <EventItem> {
Data = item
});
wasDirty = true;
}
if (wasDirty)
{
NotifySortRequested();
}
}
/* Method: ScaleTrainData
*
* Scales the inputs and outputs in the training data to the specified range.
*
* A simplified scaling method, which is mostly useful in examples where it's known that all the
* data will be in one range and it should be transformed to another range.
*
* It is not recommended to use this on subsets of data as the complete input range might not be
* available in that subset.
*
* For more powerful scaling, please consider <NeuralNet::ScaleTrain>
*
* See also:
* <ScaleOutputTrainData>, <ScaleInputTrainData>, <fann_scale_train_data at http://libfann.github.io/fann/docs/files/fann_train-h.html#fann_scale_train_data>
*
* This function appears in FANN >= 2.0.0.
*/
public void ScaleTrainData(int new_min, int new_max)
{
InternalData.scale_train_data(new_min, new_max);
}
public override string ToString()
{
StringBuilder __sb = new StringBuilder("THTokenInfo(");
bool __first = true;
if (ReferrerUserId != null && __isset.referrerUserId)
{
if (!__first)
{
__sb.Append(", ");
}
__first = false;
__sb.Append("ReferrerUserId: ");
__sb.Append(ReferrerUserId);
}
if (Token != null && __isset.token)
{
if (!__first)
{
__sb.Append(", ");
}
__first = false;
__sb.Append("Token: ");
__sb.Append(Token);
}
if (Provider != null && __isset.provider)
{
if (!__first)
{
__sb.Append(", ");
}
__first = false;
__sb.Append("Provider: ");
__sb.Append(Provider);
}
if (__isset.firstMatch)
{
if (!__first)
{
__sb.Append(", ");
}
__first = false;
__sb.Append("FirstMatch: ");
__sb.Append(FirstMatch);
}
if (LinkParams != null && __isset.linkParams)
{
if (!__first)
{
__sb.Append(", ");
}
__first = false;
__sb.Append("LinkParams: ");
__sb.Append(LinkParams.ToDebugString());
}
if (InternalData != null && __isset.internalData)
{
if (!__first)
{
__sb.Append(", ");
}
__first = false;
__sb.Append("InternalData: ");
__sb.Append(InternalData.ToDebugString());
}
if (__isset.guaranteedMatch)
{
if (!__first)
{
__sb.Append(", ");
}
__first = false;
__sb.Append("GuaranteedMatch: ");
__sb.Append(GuaranteedMatch);
}
if (OriginalData != null && __isset.originalData)
{
if (!__first)
{
__sb.Append(", ");
}
__first = false;
__sb.Append("OriginalData: ");
__sb.Append(OriginalData.ToDebugString());
}
if (Debug != null && __isset.debug)
{
if (!__first)
{
__sb.Append(", ");
}
__first = false;
__sb.Append("Debug: ");
__sb.Append(Debug);
}
__sb.Append(")");
return(__sb.ToString());
}
public void BindInternalClass ()
{
InternalData data = new InternalData ();
data.InnerData = new Data { Opacity = 1.0f };
Rectangle rectangle = new Rectangle { Opacity = 0f };
rectangle.DataContext = data;
Binding binding = new Binding ("InnerData.Opacity");
rectangle.SetBinding (Shape.OpacityProperty, binding);
}
public bool TryGetEntry(object key, out ICacheEntry entry)
{
return(InternalData.TryGetValue(key, out entry));
}
public ExtendedXlsFile TestExecuteBalansHierLev0_Valtage_220_330_Interval(BalansHierLev0Result HierLev0,
TExportExcelAdapterType ExportType, IVisualDataRequestObjectsNames getNameInterface)
{
Dictionary <ID_TypeHierarchy, string> dictionaryOfNames = HierLev0.DictionaryOfNames;
InternalData internalData = new InternalData(ExportType, HierLev0.NumbersValues);
InitBlock initBlock = new InitBlock(internalData);
Classes.TitleInfo titleData =
new Classes.TitleInfo(getNameInterface.GetBalanceNameForHierLev0(HierLev0.BalanceId), HierLev0.DTStart, HierLev0.DTEnd);
TitleBlock titleBlock = new TitleBlock(titleData);
initBlock.AddBlock(titleBlock);
List <DateTime> dateTimeListForPeriod = getNameInterface.GetDateTimeListForPeriod(HierLev0.DTStart, HierLev0.DTEnd,
HierLev0.DiscreteType, HierLev0.IsSummerOrWinter);
List <DateTimePeriod> dateTimePeriods = new List <DateTimePeriod>();
for (int i = 0; i < dateTimeListForPeriod.Count; i += 1)
{
var dateList = dateTimeListForPeriod.GetRange(i, Math.Min(2, dateTimeListForPeriod.Count - i));
if (dateList.Count > 1)
{
dateTimePeriods.Add(new DateTimePeriod(dateList[0], dateList[1]));
}
}
dateTimePeriods.Add(new DateTimePeriod(dateTimeListForPeriod.Max(), HierLev0.DTEnd.AddMinutes(30)));
HeaderFooterBlock headerFooter = new HeaderFooterBlock(new Data.HeaderFooterData(HierLev0.VoltageClass, dateTimePeriods));
initBlock.AddBlock(headerFooter);
foreach (TIntegral_HierLev0_Values balanceSection in HierLev0.Result_Values)
{
BalansHierLev0.Data.Full.BalancePartData balData = new BalansHierLev0.Data.Full.BalancePartData(getNameInterface.GetBalanceSectionName(balanceSection.HierLev0Group_Name),
HierLev0.BalPartList.Where(x => x.IsUseInGeneralBalance).Select(x => x.Name).Contains(balanceSection.HierLev0Group_Name),
HierLev0.BalPartList.Where(x => x.IsRsk).Select(x => x.Name).Contains(balanceSection.HierLev0Group_Name));
BalancePartBlock balPartBlock = new BalancePartBlock(balData);
foreach (KeyValuePair <ID_IsOurSide, TIntegral_PS_ValuesForHierLev0> psBalSect in balanceSection.HierLev0DetailGroupResult)
{
string psName;
TIntegral_PS_ValuesForHierLev0 psBalSectData = psBalSect.Value;
ID_IsOurSide side = psBalSect.Key;
ID_TypeHierarchy key = new ID_TypeHierarchy(enumTypeHierarchy.Dict_PS, -1);
key.TypeHierarchy = side.IsOurSide ? enumTypeHierarchy.Dict_PS : enumTypeHierarchy.Dict_Contr_PS;
key.ID = side.ID;
if (!dictionaryOfNames.TryGetValue(key, out psName))
{
psName = getNameInterface.GetPSName(side.ID, !side.IsOurSide);
}
PsBlock psBlock = new PsBlock(psName);
foreach (TI_Integral_ValuesForHierLev0 tiPsBalSect in psBalSectData.TI_List)
{
string tIName = string.Empty;
key.TypeHierarchy = tiPsBalSect.TypeHierarchy;
key.ID = tiPsBalSect.ID;
if (!dictionaryOfNames.TryGetValue(key, out tIName))
{
switch (tiPsBalSect.TypeHierarchy)
{
case enumTypeHierarchy.Dict_PS:
tIName = getNameInterface.GetPSName(tiPsBalSect.ID, !side.IsOurSide);
break;
case enumTypeHierarchy.Info_TI:
tIName = getNameInterface.GetTIName(tiPsBalSect.ID, false);
break;
case enumTypeHierarchy.Info_ContrTI:
tIName = getNameInterface.GetTIName(tiPsBalSect.ID, true);
break;
case enumTypeHierarchy.Info_TP:
tIName = getNameInterface.GetTPName(tiPsBalSect.ID);
break;
}
}
List <TVALUES_DB> inputValues;
List <TVALUES_DB> outputValues;
Data.TiData tiData = new Data.TiData(tIName);
if (tiPsBalSect.Val_List.TryGetValue(1, out inputValues))
{
tiData.InputInterval = inputValues.Select(x => new TVALUES_DB(x.F_FLAG, x.F_VALUE / 1000)).ToList();
}
if (tiPsBalSect.Val_List.TryGetValue(2, out outputValues))
{
tiData.OutputInterval = outputValues.Select(x => new TVALUES_DB(x.F_FLAG, x.F_VALUE / 1000)).ToList();
}
TiBlock tiBlock = new TiBlock(tiData);
psBlock.AddBlock(tiBlock);
}
balPartBlock.AddBlock(psBlock);
}
headerFooter.AddBlock(balPartBlock);
}
ExtendedXlsFile xls = new ExtendedXlsFile(ExportType);
initBlock.Render(xls);
return(xls);
}
/* Method: SaveTrain
*
* Save the training structure to a file, with the format as specified in <ReadTrainFromFile>
*
* Return:
* The function returns true on success and false on failure.
*
* See also:
* <ReadTrainFromFile>, <SaveTrainToFixed>, <fann_save_train at http://libfann.github.io/fann/docs/files/fann_train-h.html#fann_save_train>
*
* This function appears in FANN >= 1.0.0.
*/
public bool SaveTrain(string filename)
{
return(InternalData.save_train(filename));
}
/* Method: Dispose
*
* Disposes of the training data.
*/
public void Dispose()
{
InternalData.Dispose();
}
/* Method: SaveTrainToFixed
*
* Saves the training structure to a fixed point data file.
*
* This function is very useful for testing the quality of a fixed point network.
*
* Return:
* The function returns true on success and false on failure.
*
* See also:
* <SaveTrain>, <fann_save_train_to_fixed at http://libfann.github.io/fann/docs/files/fann_train-h.html#fann_save_train_to_fixed>
*
* This function appears in FANN >= 1.0.0.
*/
public bool SaveTrainToFixed(string filename, uint decimalPoint)
{
return(InternalData.save_train_to_fixed(filename, decimalPoint));
}
protected bool Equals(InternalData other)
{
return(CVBucketed.Equals(other.CVBucketed) && Area.Equals(other.Area));
}
/// <summary>
/// チェック用データをキューに入れます。
/// </summary>
private static void PushData(InternalData data, bool highPriority)
{
if (data == null)
{
return;
}
lock (targetList)
{
if (highPriority)
{
targetList.Insert(0, data);
}
else
{
targetList.Add(data);
}
Monitor.PulseAll(targetList);
}
}
public void Calculator_Calculate_OneDoctorWithTwoPacientsAndMeasurementsForOnePacient()
{
var data = new InternalData();
var doc = new InternalDataDoctor() { DoctorId = 1 };
var pac0 = new InternalDataPacient() { DoctorId = doc.DoctorId, PacientId = 1 };
var pac1 = new InternalDataPacient() { DoctorId = doc.DoctorId, PacientId = 2 };
var me00 = new InternalDataMeasurement() { PacientId = pac0.PacientId, MeasurementId = 1 };
var me01 = new InternalDataMeasurement() { PacientId = pac0.PacientId, MeasurementId = 2 };
data.Doctors.Add(doc);
data.Pacients.Add(pac0);
data.Measurements.Add(me00);
data.Measurements.Add(me01);
var res0 = Calc.Calculate(data, CalculationType.ByDoctor, excludeDoctorsWithoutPacients: false, excludePacientsWithoutMeasurements: true);
//pacient 0 with bad measurement (result should equals 0)
Assert.AreEqual(1, res0.Count(), "res0 must contains 1 result");
Assert.AreEqual(0d, res0.First().Result, "result for res0 must equals 0");
var hc00 = new InternalDataHeighComponent() { MeasurementId = me00.MeasurementId, Height = 100 };
var hc01 = new InternalDataHeighComponent() { MeasurementId = me01.MeasurementId, Height = 200 };
var dt = DateTime.Now;
var ts00 = new InternalDataTimestamp() { MeasurementId = me00.MeasurementId, Timestamp = dt };
var ts01 = new InternalDataTimestamp() { MeasurementId = me01.MeasurementId, Timestamp = dt.AddDays(7) };
data.HeighComponent.Add(hc00);
data.HeighComponent.Add(hc01);
data.Timestamps.Add(ts00);
data.Timestamps.Add(ts01);
var res1 = Calc.Calculate(data, CalculationType.ByDoctor, excludeDoctorsWithoutPacients: false, excludePacientsWithoutMeasurements: true);
//pacient 0 with bad measurement (result should equals 0)
Assert.AreEqual(1, res1.Count(), "res1 must contains 1 result");
Assert.AreEqual(100d, res1.First().Result, "result for res1 must equals 100");
//add one pacient without measurement
data.Pacients.Add(pac1);
var res2 = Calc.Calculate(data, CalculationType.ByDoctor, excludeDoctorsWithoutPacients: false, excludePacientsWithoutMeasurements: true);
Assert.AreEqual(1, res2.Count(), "res2 must contains 1 result");
Assert.AreEqual(100d, res2.First().Result, "result for res2 must equals 100");
//if we calc 2 pacients (one without measurements) then result must be (100 - 0) / 2 = 50
var res3 = Calc.Calculate(data, CalculationType.ByDoctor, excludeDoctorsWithoutPacients: false, excludePacientsWithoutMeasurements: false);
Assert.AreEqual(1, res3.Count(), "res3 must contains 1 result");
Assert.AreEqual(50d, res3.First().Result, "result for res3 must equals 50");
}
private IEnumerator Tick(InternalData internalData)
{
SendInternalRpc(internalData.remoteId, InternalMethod.Tock);
yield return(null);
}
/* Method: ShuffleTrainData
*
* Shuffles training data, randomizing the order.
* This is recommended for incremental training, while it have no influence during batch training.
*
* This function appears in FANN >= 1.1.0.
*/
public void ShuffleTrainData()
{
InternalData.shuffle_train_data();
}
private void Controller_ResizeEnd(object sender, EventArgs e)
{
InternalData.SettingsGrid.Height = Height;
InternalData.SettingsGrid.Width = Width;
InternalData.SaveSettings();
}
public void FetchData(DataStore.MainInterface dataStore)
{
if (dataStore == null)
throw new ArgumentNullException("dataStore");
var d = new InternalData();
d.Doctors = dataStore.Entities
.Where(e => IsDoctor(e))
.Select(e => new InternalDataDoctor() { DoctorId = e.Id })
.ToList();
d.Pacients = dataStore.Relations
.Where(r => IsDoctor(r.Item1) && IsPacient(r.Item2))
.Select(r => new InternalDataPacient() { DoctorId = r.Item1.Id, PacientId = r.Item2.Id })
.ToList();
d.Measurements = dataStore.Relations
.Where(r => IsPacient(r.Item1) && IsMeasurement(r.Item2))
.Select(r => new InternalDataMeasurement() { PacientId = r.Item1.Id, MeasurementId = r.Item2.Id })
.ToList();
//var dmIds = d.Measurements.Select(m => m.MeasurementId).ToArray();
var dmIds = d.Measurements.Select(m => m.MeasurementId).ToDictionary(m => m);
d.Timestamps = dataStore.TimestampComponents
.Where(ts => !ts.IsMissing)
//.Join(dmIds, tc => tc.EntityId, i => i, (ts, i) => ts)
.Where(ts => dmIds.ContainsKey(ts.EntityId))
.Select(ts => new InternalDataTimestamp()
{
MeasurementId = ts.EntityId,
Timestamp = ts.Timestamp
})
.ToArray();
d.HeighComponent = dataStore.HeightComponents
.Where(h => !h.IsMissing)
//.Join(dmIds, hc => hc.EntityId, i => i, (hs, i) => hs)
.Where(h => dmIds.ContainsKey(h.EntityId))
.Select(h => new InternalDataHeighComponent()
{
MeasurementId = h.EntityId,
Height = (h.Unit == DataStore.LengthUnit.Inch ? h.Value : h.Value * 39.37) //Meters to inches
})
.ToArray();
data = d;
}
/* Method: MergeTrainData
*
* Merges the data into the data contained in the <TrainingData>.
*
* This function appears in FANN >= 1.1.0.
*/
public void MergeTrainData(TrainingData data)
{
InternalData.merge_train_data(data.InternalData);
}
/// <summary>
/// 生放送に必要な情報を非同期でまとめて取得します。
/// </summary>
public static LiveStreamInfo GetLiveStreamInfo(PlayerStatus playerStatus,
CookieContainer cc)
{
if (playerStatus == null)
{
throw new ArgumentNullException("playerStatus");
}
if (cc == null)
{
throw new ArgumentNullException("cc");
}
// 結果などを保存するオブジェクトです。
var internalData = new InternalData()
{
Cookie = cc,
};
var eventList = new AutoResetEvent[3]
{
new AutoResetEvent(false),
new AutoResetEvent(false),
new AutoResetEvent(false),
};
// playerStatusを取得します。
internalData.LiveStreamInfo.PlayerStatus = playerStatus;
eventList[0].Set();
// publishstatusを取得します。
// これは放送主にしか取得できません。
if (playerStatus.Stream.IsOwner)
{
BeginGetPublishStatus(
playerStatus.Stream.Id,
internalData,
eventList[1]);
}
else
{
eventList[1].Set();
}
// 放送ページの情報を取得します。
// playerstatusなどでは放送タイトルなどが一定の文字数で
// 省略されてしまうためです。
BeginGetLiveInfo(
playerStatus.Stream.Id,
internalData,
eventList[2]);
// タイムアウト時間を取得します。
// (デフォルトで３０秒)
var timeout = WebUtil.DefaultTimeout;
if (timeout < 0)
{
timeout = 30 * 1000;
}
// 各イベントが終了するのを待ちます。
foreach (var ev in eventList)
{
if (!ev.WaitOne(TimeSpan.FromMilliseconds(timeout)))
{
throw new TimeoutException(
"放送情報の取得がタイムアウトしました。");
}
// 例外があればそれをここで投げ返します。
var ex = internalData.Exception;
if (ex != null)
{
throw new NicoLiveException(
"GetLiveStreamInfoでエラーが発生しました。", ex);
}
}
return internalData.LiveStreamInfo;
}
/* Method: GetTrainInput
* Gets the training input data at the given position
*
* Returns:
* An array of input training data at the given position
*
* See also:
* <GetTrainOutput>, <SetTrainData>
*
* This function appears in FANN >= 2.3.0.
*/
public DataAccessor GetTrainInput(uint position)
{
DataAccessor data = DataAccessor.FromPointer(InternalData.get_train_input(position), (int)InputCount);
return(data);
}
/// <summary>
/// publishstatusの取得を開始します。
/// </summary>
private static void BeginGetPublishStatus(long liveId,
InternalData internalData,
AutoResetEvent ev)
{
WebUtil.RequestHttpAsync(
NicoString.GetPublishStatusUrl(liveId),
null,
internalData.Cookie,
(result, data) =>
{
try
{
// publishstatusは放送主でないと取得できません。
var v = PublishStatus.CreateFromXml(liveId, GetXml(data));
internalData.LiveStreamInfo.PublishStatus = v;
}
catch (Exception ex)
{
internalData.Exception = ex;
}
ev.Set();
});
}
/* Method: GetTrainOutput
* Gets the training output data at the given position
*
* Returns:
* An array of output training data at the given position
*
* See also:
* <GetTrainInput>
*
* This function appears in FANN >= 2.3.0.
*/
public DataAccessor GetTrainOutput(uint position)
{
return(DataAccessor.FromPointer(InternalData.get_train_output(position), (int)OutputCount));
}
/* Method: ScaleTrainData
*
* Scales the inputs and outputs in the training data to the specified range.
*
* A simplified scaling method, which is mostly useful in examples where it's known that all the
* data will be in one range and it should be transformed to another range.
*
* It is not recommended to use this on subsets of data as the complete input range might not be
* available in that subset.
*
* For more powerful scaling, please consider <NeuralNet::ScaleTrain>
*
* See also:
* <ScaleOutputTrainData>, <ScaleInputTrainData>, <fann_scale_train_data at http://libfann.github.io/fann/docs/files/fann_train-h.html#fann_scale_train_data>
*
* This function appears in FANN >= 2.0.0.
*/
public void ScaleTrainData(double new_min, double new_max)
{
InternalData.scale_train_data(new_min, new_max);
}
/* Method: SubsetTrainData
*
* Changes the training data to a subset, starting at position *pos*
* and *length* elements forward. Use the copy constructor to work
* on a new copy of the training data.
*
* >TrainingData fullDataSet = new TrainingData();
* >fullDataSet.ReadTrainFromFile("somefile.train");
* >TrainingData smallDataSet = new TrainingData(fullDataSet);
* >smallDataSet->SubsetTrainData(0, 2); // Only use first two
* >// Use smallDataSet ...
* >small_data_set.Dispose();
*
* See also:
* <fann_subset_train_data http://libfann.github.io/fann/docs/files/fann_train-h.html#fann_subset_train_data>
*
* This function appears in FANN >= 2.0.0.
*/
public void SubsetTrainData(uint pos, uint length)
{
InternalData.subset_train_data(pos, length);
}
/// <summary>
/// Calculate stored data
/// </summary>
/// <param name="data">Internal data storage</param>
/// <param name="calcType">Calculate average data by each doctor or by each doctor's pacient</param>
/// <param name="calcTimeType">Calculate results by any part of year (year/month/week/day)</param>
/// <param name="excludeDoctorsWithoutPacients">Excludes from result doctors with no one pacient</param>
/// <param name="excludePacientsWithoutMeasurements">Excludes from result pacients with one or no one measuremtns</param>
/// <returns>Calculation result</returns>
public IEnumerable<AnalysisModule.SimpleCalculationResult> Calculate(
InternalData data,
CalculationType calcType = CalculationType.ByDoctor,
CalculationTimeType calcTimeType = CalculationTimeType.PerWeek,
bool excludeDoctorsWithoutPacients = false,
bool excludePacientsWithoutMeasurements = false)
{
if (data == null)
throw new ArgumentNullException("data");
if (data.Doctors == null)
throw new ArgumentNullException("data.Doctors", "Feel 'Doctors' array before calling Calculation");
if (data.Pacients == null)
throw new ArgumentNullException("data.Pacients", "Feel 'Pacients' array before calling Calculation");
if (data.Timestamps == null)
throw new ArgumentNullException("data.Timestamps", "Feel 'Timestamps' array before calling Calculation");
if (data.Measurements == null)
throw new ArgumentNullException("data.Measurements", "Feel 'Measurements' array before calling Calculation");
if (data.HeighComponent == null)
throw new ArgumentNullException("data.HeighComponent", "Feel 'HeighComponent' array before calling Calculation");
if (!data.Doctors.Any())
throw new ArgumentException("You must feel doctors array before calling Calculation", "data.Doctors");
var res = new List<AnalysisModule.SimpleCalculationResult>();
var bigTableData =
data.Doctors
.LeftOuterJoin(data.Pacients, d => d.DoctorId, p => p.DoctorId, (d, p) => new { DoctorId = d.DoctorId, PacientId = p == null ? (long?)null : p.PacientId })
.LeftOuterJoin(data.Measurements, i => i.PacientId, m => m.PacientId, (i, m) => new { i.DoctorId, i.PacientId, MeasurementId = m == null ? (long?)null : m.MeasurementId })
.LeftOuterJoin(data.Timestamps, i => i.MeasurementId, t => t.MeasurementId, (i, t) => new { i.DoctorId, i.PacientId, i.MeasurementId, Timestamp = t == null ? (DateTime?)null : t.Timestamp })
.LeftOuterJoin(data.HeighComponent
.Where(h => h.Height > 0
&& !double.IsInfinity(h.Height)
&& !double.IsNegativeInfinity(h.Height)
&& !double.IsPositiveInfinity(h.Height)
&& !double.IsNaN(h.Height)
),
i => i.MeasurementId, hc => hc.MeasurementId, (i, hc) => new { i.DoctorId, i.PacientId, i.MeasurementId, i.Timestamp, Height = hc == null ? (double?)null : hc.Height })
.ToArray();
//Table generated. Group data by Doctor, Pacient and Measurement
var doctors = bigTableData
.GroupBy(i => i.DoctorId)
.Select(g => new
{
g.FirstOrDefault().DoctorId,
//Group data by pacient
Pacients = g.Where(p => p != null)
.GroupBy(g2 => g2.PacientId)
.Select(g2 => new
{
g2.FirstOrDefault().PacientId,
Measurements = g2
.Where(i => i.Height != null && i.Timestamp != null)
.OrderBy(m => m.Timestamp)
.ToArray()
})
.Select(g2 => new
{
g2.PacientId,
MeasurementsExists = g2.Measurements.Count() > 1,
Measurements = g2.Measurements.Count() > 1
? Enumerable.Range(0, g2.Measurements.Count() - 1)
.Select(i => new //Get measurement pairs for pacient
{
Start = g2.Measurements.ElementAt(i),
End = g2.Measurements.ElementAt(i + 1)
})
.Select(i => new //Calc start and end measure data
{
HeightStart = i.Start.Height.Value,
HeightEnd = i.End.Height.Value,
TimestampStart = i.Start.Timestamp.Value,
TimestampEnd = i.End.Timestamp.Value,
})
.Select(i => new //get height change by part of time
{
HeightChange = i.HeightEnd - i.HeightStart,
TimeParts = GetTimePart((i.TimestampEnd - i.TimestampStart).TotalDays, calcTimeType),
})
.Select(i => i.HeightChange / i.TimeParts) //calc growth per week
.ToArray()
: Enumerable.Empty<double>().ToArray() //return empty array if no one or one only measurement
})
.Select(g2 => new
{
g2.PacientId,
g2.MeasurementsExists,
//g2.Measurements //!!! enable to debug
MeanRateGrowhPerTimePart = g2.MeasurementsExists ? g2.Measurements.Average() : 0
})
})
//Exclude from data pacients with one or less measurements (is setted)
.Select(i => new
{
i.DoctorId,
Pacients = i.Pacients
.Where(p => p.PacientId != null)
.Where(n => !excludePacientsWithoutMeasurements || n.MeasurementsExists)
.ToArray()
})
//Exclude from data doctors without pacients (is setted)
.Where(i => !excludeDoctorsWithoutPacients || i.Pacients.Any())
.ToArray()
;
//Get enum description attribute
var perName = calcTimeType.ToString();
var memInfo = calcTimeType.GetType().GetMember(perName).FirstOrDefault();
if (memInfo != null)
{
var descrAttr = memInfo.GetCustomAttributes(typeof(DescriptionAttribute), false).FirstOrDefault() as DescriptionAttribute;
if (descrAttr != null)
perName = "per " + descrAttr.Description;
}
//generate result from data
foreach (var doc in doctors)
{
switch (calcType)
{
case CalculationType.ByPacient:
res.AddRange(
doc.Pacients.Select(i => new AnalysisModule.SimpleCalculationResult()
{
Result = i.MeanRateGrowhPerTimePart,
AnalysisName = string.Format("{0} {3} for pacient (id:{1}) of doctor (id:{2})", CalcTaskName, i.PacientId, doc.DoctorId, perName)
}));
break;
case CalculationType.ByDoctor:
res.Add(new AnalysisModule.SimpleCalculationResult()
{
Result = doc.Pacients.Any() ? doc.Pacients.Average(p => p.MeanRateGrowhPerTimePart) : 0,
AnalysisName = string.Format("{0} {2} for doctor (id:{1})", CalcTaskName, doc.DoctorId, perName)
});
break;
}
}
return res;
}
/* Method: ScaleOutputTrainData
*
* Scales the outputs in the training data to the specified range.
*
* A simplified scaling method, which is mostly useful in examples where it's known that all the
* data will be in one range and it should be transformed to another range.
*
* It is not recommended to use this on subsets of data as the complete input range might not be
* available in that subset.
*
* For more powerful scaling, please consider <NeuralNet::ScaleTrain>
*
* See also:
* <ScaleInputTrainData>, <ScaleTrainData>, <fann_scale_output_train_data at http://libfann.github.io/fann/docs/files/fann_train-h.html#fann_scale_output_train_data>
*
* This function appears in FANN >= 2.0.0.
*/
public void ScaleOutputTrainData(float new_min, float new_max)
{
InternalData.scale_output_train_data(new_min, new_max);
}