public DatePlus(MultipleTimeseries data, bool from = true)
{
if (from)
{
mDate = mDate = new DateTime(data.getStartDate().Ticks);
}
else
{
mDate = mDate = new DateTime(data.getEndDate().Ticks);
}
mUnit = data.mTimeUnit;
mZone = TimeZoneInfo.Utc;
}
public Graph(MultipleTimeseries dataSet, Size canvasSize, String title, String titleX, String titleY, bool monthlyGraph = false, bool plotAccumulative = false)
{
this.mDataset = dataSet;
this.mPlotAccumulative = plotAccumulative;
this.mMonthlyGraph = monthlyGraph;
this.mMinX = DateTime.Now;
this.mMaxX = DateTime.Now;
this.mMinY = 0;
this.mMaxY = 0;
this.mGraphItems = new List <GraphItem>();
this.mCanvasSize = canvasSize;
this.mTitle = title;
this.mTitleX = titleX;
this.mTitleY = titleY;
this.mGraphOffset = new Point((int)Math.Max(50, Math.Min(200, mCanvasSize.Width * 0.1f)), (int)Math.Max(50, Math.Min(200, mCanvasSize.Height * 0.1f)));
Initialise();
}
public List <float[]> getTimeseries(DateTime from, DateTime to, ExtractionPoint[] points, TimeUnit unit)
{
TimePeriod period = new TimePeriod();
period.from = from;
period.to = to;
period.zone = TimeZoneInfo.Utc;
period = stripPrecision(period, unit);
int steps = MultipleTimeseries.CalcSteps(period, unit);
List <float[]> dataList = new List <float[]>();
for (int j = 0; j < points.Length; ++j)
{
float[] data = new float[steps];
DateTime iter = from;
for (int i = 0; i < steps; ++i)
{
if (mFilenameLookup.ContainsKey(iter))
{
data[i] = mFilenameLookup[iter].getData(points[j].longitude, points[j].latitude);
}
else
{
data[i] = -9999.9f;
}
iter = MultipleTimeseries.incrementDateTime(iter, unit, 1);
}
dataList.Add(data);
}
return(dataList);
}
public void Extract(String outDir)
{
// Get data for longitude and latitude
List <float[]> queryData = mGriddedDataset.getTimeseries(mExtractionPoints, mTimeUnit);
// Check dataset returned for each station
if (queryData.Count == mExtractionPoints.Length)
{
// Create data set
mMultipleTimeseries = new MultipleTimeseries(mGriddedDataset.getFromDate(), mGriddedDataset.getToDate(), mTimeUnit);
// Scale values
for (int i = 0; i < queryData.Count; ++i)
{
for (int j = 0; j < queryData[i].Length; ++j)
{
queryData[i][j] *= (float)mExtractionPoints[i].weight;
}
}
// Find unique stations
string[] uniqueStations = mExtractionPoints.Select(v => v.siteName).Distinct().ToArray();
string[] siteNames = mExtractionPoints.Select(v => v.siteName).ToArray();
int nextIndex;
foreach (String s in uniqueStations)
{
// Find first occurance
int index = Array.IndexOf(siteNames, s);
// Add all other indices and compute normalise factor
nextIndex = index;
float[] normalise = new float[queryData[index].Length]; // unique normalise value for each data point in case of missing values
for (int i = 0; i < queryData[index].Length; ++i)
{
if (queryData[index][i] < 0) // if first station value is invalid, set it to zero instead of missing and set normalise to zero
{
queryData[index][i] = 0;
normalise[i] = 0;
}
else // otherwise begin the normalise constant
{
normalise[i] = (float)mExtractionPoints[index].weight;
}
}
// Find next occurance
nextIndex = Array.IndexOf(siteNames, s, nextIndex + 1);
while (nextIndex != -1)
{
// Combine data sources for each point and compute normalisation
for (int i = 0; i < queryData[index].Length; ++i)
{
// if valid data point, add it to the normalise value
if (queryData[nextIndex][i] >= 0)
{
queryData[index][i] += queryData[nextIndex][i];
normalise[i] += (float)mExtractionPoints[nextIndex].weight;
}
}
// Find next occurance
nextIndex = Array.IndexOf(siteNames, s, nextIndex + 1);
}
// Normalise data
for (int i = 0; i < queryData[index].Length; ++i)
{
if (normalise[i] > 0)
{
queryData[index][i] /= normalise[i];
}
else
{
queryData[index][i] = -9999.9f;
}
}
// Add to data set
mMultipleTimeseries.AddColumn(new SingleTimeseries(s, queryData[index]));
}
// Save data
mMultipleTimeseries.writeToFile(outDir + Path.DirectorySeparatorChar + "Data.csv");
}
}
