private void testWithLines(StringBuilder sb)
{
var reader = new StreamReader(new MemoryStream(Encoding.UTF8.GetBytes(sb.ToString())));
Predict.doPredict(reader, writer, testModel);
}
private int selected; //for the pick state
// Use this for initialization
void Start()
{
canvas.SetActive(false);
normalState.SetActive(true);
drawingState.SetActive(false);
pickState.SetActive(false);
wrongState.SetActive(false);
editState.SetActive(false);
systemState.SetActive(false);
canvasOpen = false;
predictor = GameObject.FindGameObjectWithTag("Predictor").GetComponent <Predict> ();
selector = GameObject.FindGameObjectWithTag("ScaleSelection").GetComponent <Selection> ();
scaler = GameObject.FindGameObjectWithTag("ScaleSelection").GetComponent <ScaleScript> ();
mapping = new int[10];
mapping [0] = 4;
mapping [1] = 6;
mapping [2] = 5;
mapping [3] = 8;
mapping [4] = 10;
mapping [5] = 1;
mapping [6] = 2;
mapping [7] = 3;
mapping [8] = 7;
mapping [9] = 9;
}
public static List <StockControlModelPredict> GetPredictModel(int beverageId)
{
DateTime lastWeek = DateTime.Today.AddDays(-7);
using (var context = new KoreanStudyCafeEntities())
{
var query = from x in context.BeverageRecords
orderby x.DayQuarter
where x.BeverageID == beverageId && x.Date == lastWeek
select x;
List <BeverageRecord> list = query.ToList();
var query2 = from x in list
select new StockControlModelPredict
{
UserCount = x.UserCount,
DayQuater = x.DayQuarter,
Usage = Predict.Predictor(beverageId, x.DayQuarter, x.UserCount, GetDay(DateTime.Now))
};
return(query2.ToList());
//return list.ConvertAll(x => new StockConreolModelPredictCoffee
//{
// UserCount = x.UserCount,
// DayQuater = x.DayQuarter,
// Usage = Predict.Predictor(1, x.DayQuarter, x.UserCount, GetDay(DateTime.Now))
//});
}
}
static void Main(string[] args)
{
string[] files = Directory.GetFiles(@"C:\Users\brush\Documents\Visual Studio 2012\Projects\SetSpotter\SetSpotter\testimages\");
int i = 0;
Predict red = new Predict(@"C:\Users\brush\Desktop\round2\red.txt.model");
Predict purple = new Predict(@"C:\Users\brush\Desktop\round2\purple.txt.model");
Predict green = new Predict(@"C:\Users\brush\Desktop\round2\green.txt.model");
//foreach (string file in files)
{
FoundColorSpaces foundColorSpaces = ColorSpaceFinder.Find(@"C:\Users\brush\Desktop\IMAG2112.jpg");
FoundBlobs foundBlobs = BlobFinder.Find(foundColorSpaces, 80, 25, 90, 50, 1.2, 2.2);
foreach (Blob blob in foundBlobs.Blobs)
{
//Bitmap color = foundColorSpaces.OriginalColorSpace.Clone(blob.Rectangle, PixelFormat.Format24bppRgb);
Bitmap color = ColorSpaceFinder.FindColorCorrectedForBlob(foundColorSpaces, blob);
ColorTypeEnum colorType = ColorSpaceFinder.FindShapeColor(color, red, purple, green);
using (Graphics g = Graphics.FromImage(color))
{
g.DrawString(colorType.ToString(), new Font("Arial", 12), Brushes.Black, new PointF(0, 5));
}
color.Save(@"C:\users\brush\desktop\blobs\" + (i++).ToString() + ".bmp");
}
}
return;
}
public Record getPredict(string ctsMarker, ClientObjectAttribute cl)
{
Program dbscan = new Program();
//Debug.Log("!!!!!!!!!!!!!Start Dbscan!!!!!!!!!!!!!");
List <int> clusterIds;
List <Cluster> clusterDes;
HashSet <Record[]> _clusters = dbscan.Dbscan(out clusterIds, out clusterDes);
Predict predict = new Predict();
List <Record> gazePoints = preGaze.getGazePoints(getAllBrowseRecord());
List <Record> predictedCluster = predict.GetPredictedCluster(
new Vector3(cl.CameraPosX, cl.CameraPosY, cl.CameraPosZ),
new Vector3(cl.CameraRotX, cl.CameraRotY, cl.CameraRotZ),
_clusters, clusterDes
);
Record posRange = getRangeRecord(predictedCluster);
Vector3 tmpPos = pathHistory[ctsMarker].cameraPos;
Vector3 rot;
float x_sum = 0, y_sum = 0, z_sum = 0, x_rot = 0, y_rot = 0, z_rot = 0;
foreach (var j in predictedCluster)
{
x_sum += (float)j.posX;
y_sum += (float)j.posY;
z_sum += (float)j.posZ;
x_rot += (float)j.rotX;
y_rot += (float)j.rotY;
z_rot += (float)j.rotZ;
}
float x_ave = x_sum / predictedCluster.Count;
float y_ave = y_sum / predictedCluster.Count;
float z_ave = z_sum / predictedCluster.Count;
float x_averot = x_rot / predictedCluster.Count;
float y_averot = y_rot / predictedCluster.Count;
float z_averot = z_rot / predictedCluster.Count;
if (tmpPos.x <= posRange.posX && tmpPos.y <= posRange.posY && tmpPos.z <= posRange.posZ &&
tmpPos.x >= posRange.rotX && tmpPos.y >= posRange.rotY && tmpPos.z >= posRange.rotZ)
{
rot = new Vector3(x_averot, y_averot, z_averot);
}
else
{
rot = new Vector3(x_ave, y_ave, z_ave);
rot = rot - tmpPos;
}
//------
/*if (predictedCluster != null)
* {
* //把块变成坐标系
* var exchangeAxis = new ExchangeAxis();
* Record pos = new Record(exchangeAxis.ModelIndex_to_unityPos(x_ave, y_ave, z_ave), new Vector3(x_averot, y_averot, z_averot));
* return pos;
*
* }
* else { Debug.Log("No suggestion"); }*/
return(new Record(MoveForwardPredict(pathHistory[ctsMarker], rot), rot));
}
public Record getPredict(string ctsMarker, ClientObjectAttribute cl)
{
Debug.Log("Start predict");
Predict predict = new Predict();
List <Record> gazePoints = preGaze.getGazePoints(getAllBrowseRecord());
Cluster predictedCluster = predict.GetPredictedCluster(
new Vector3(cl.CameraPosX, cl.CameraPosY, cl.CameraPosZ),
new Vector3(cl.CameraRotX, cl.CameraRotY, cl.CameraRotZ),
clusterNet, gazePoints);
//Record posRange = getRangeRecord(predictedCluster);
Vector3 tmpPos = pathHistory[ctsMarker].cameraPos;
//float x_sum = 0, y_sum = 0, z_sum = 0, x_rot = 0, y_rot = 0, z_rot = 0;
//foreach (var j in predictedCluster)
//{
// x_sum += (float)j.posX;
// y_sum += (float)j.posY;
// z_sum += (float)j.posZ;
// x_rot += (float)j.rotX;
// y_rot += (float)j.rotY;
// z_rot += (float)j.rotZ;
//}
//float x_ave = x_sum / predictedCluster.Count;
//float y_ave = y_sum / predictedCluster.Count;
//float z_ave = z_sum / predictedCluster.Count;
Vector3 cluterPos = predictedCluster.center;
//float x_averot = x_rot / predictedCluster.Count;
//float y_averot = y_rot / predictedCluster.Count;
//float z_averot = z_rot / predictedCluster.Count;
Vector3 rot = cluterPos - tmpPos;
//if (tmpPos.x <= posRange.posX && tmpPos.y <= posRange.posY && tmpPos.z <= posRange.posZ
// && tmpPos.x >= posRange.rotX && tmpPos.y >= posRange.rotY && tmpPos.z >= posRange.rotZ)
//{
// rot = new Vector3(x_averot, y_averot, z_averot);
//}
//else
//{
// rot = new Vector3(x_ave, y_ave, z_ave);
// rot = rot - tmpPos;
//}
//------
/*if (predictedCluster != null)
* {
* //把块变成坐标系
* var exchangeAxis = new ExchangeAxis();
* Record pos = new Record(exchangeAxis.ModelIndex_to_unityPos(x_ave, y_ave, z_ave), new Vector3(x_averot, y_averot, z_averot));
* return pos;
*
* }
* else { Debug.Log("No suggestion"); }*/
return(new Record(MoveForwardPredict(pathHistory[ctsMarker], rot), rot));
}
private void Init()
{
isFirstLoad = true;
Predictions = null;
python = !string.IsNullOrEmpty(pythonPath) ? new PythonExecutor(pythonPath) : new PythonExecutor();
python.AddStandartOutputErrorFilters("Using TensorFlow backend.");
python.AddStandartOutputErrorFilters("CUDA_ERROR_NO_DEVICE: no CUDA-capable device is detected");
client = new PipeClient();
predict = new Predict(client);
python.OnPythonError += Python_OnPythonError;
TinfoBox.Start(initText, title);
python.RunScript("main.py");
client.Connect("openstsm");
}
void Brain.IObserver.onPredict(Predict response)
{
if (response?.prediction == null)
{
_logger.warning($"Received null prediction from server...");
return;
}
var snap = new BodyVectorFloat();
for (int i = 0; i < snap.length; ++i)
{
snap[i] = response.prediction[i];
}
_logger.info($"Received prediction from server: {_targetSnapshot}");
_targetSnapshot = snap;
}
protected void Page_Load(object sender, EventArgs e)
{
ServiceReference1.Service1Client service = new ServiceReference1.Service1Client();
ServiceReference1.UserDetail user = (ServiceReference1.UserDetail)HttpContext.Current.Session["User"];
if (user.AccessLvl == 1)
{
ServiceReference1.reportDonation[] donations = service.getAllDailyDonations(service.getNPOWithManager(user.userValue.UserID));
ServiceReference1.reportDonation[] don = Predict.prodictValue(donations, 6);
dates = new string[don.Count()];
amounts = new double[don.Count()];
for (int i = 0; i < don.Count(); i++)
{
dates[i] = don[i].Date;
amounts[i] = don[i].Amount;
}
}
}
public void RunSelectiveSearch()
{
Predict predict;
PythonExecutor python = new PythonExecutor(@"C:\Users\kutiatore\AppData\Local\Programs\Python\Python35\python.exe");
python.AddStandartOutputErrorFilters("Using TensorFlow backend.");
python.AddStandartOutputErrorFilters("CUDA_ERROR_NO_DEVICE: no CUDA-capable device is detected");
PipeClient client = new PipeClient();
predict = new Predict(client);
try
{
python.RunScript("main.py");
client.Connect("openstsm");
predict = new Predict(client);
bool run = predict.LoadModel(@"E:\\Storage\\Python\\OpenSTSM\\ML\\models\\model.model");
run = predict.ImageDimessionCorrections("E:\\Libraries\\Desktop\\Visa Docs\\test_selective_1.png");
run = predict.RunSelectiveSearch(80, 1.0f);
string results = predict.RunPrediction(5, 3, 5, 1, 8, 2, true);
client.Close();
python.Close();
}
catch (System.Exception e)
{
System.Diagnostics.Debug.WriteLine("---------------------");
System.Diagnostics.Debug.WriteLine(e.StackTrace);
System.Diagnostics.Debug.WriteLine("---------------------");
if (client.isConnected())
{
client.Close();
}
python.Close();
}
}
public VersionController(Predict predict, ITextClassifier textClassifier)
{
_predict = predict;
_textClassifier = textClassifier;
}
public static async Task <Predict> GetImgUrlPrediction(string accessToken, List <string> imgURLs)
{
ServicePointManager.ServerCertificateValidationCallback += new RemoteCertificateValidationCallback(ValidateServerCertificate);
//
// MESSAGE CONTENT
//string postData = "client_id="+ clientId + "&client_secret="+ clientSecret + "&grant_type=client_credentials";
List <PredictInput> inputs = new List <PredictInput>();
foreach (string imgUrl in imgURLs)
{
inputs.Add(new PredictInput
{
Data = new PredictImage
{
Image = new PredictImageData
{
Url = imgUrl
}
}
});
}
var ins = new
{
Inputs = inputs
};
string postData = LowercaseJsonSerializer.SerializeObject(ins);
byte[] byteArray = Encoding.UTF8.GetBytes(postData);
//
// CREATE REQUEST
HttpWebRequest Request = (HttpWebRequest)WebRequest.Create(_apiEndPoint + _predictPath);
Request.Method = "POST";
Request.KeepAlive = false;
Request.ContentType = "application/json";
Request.Headers.Add("cache-control", "no-cache");
Request.Headers.Add("authorization", "Bearer " + accessToken);
Stream dataStream = await Request.GetRequestStreamAsync();
await dataStream.WriteAsync(byteArray, 0, byteArray.Length);
dataStream.Close();
//
// SEND MESSAGE
try
{
WebResponse Response = await Request.GetResponseAsync();
StreamReader Reader = new StreamReader(Response.GetResponseStream());
string responseLine = await Reader.ReadToEndAsync();
Reader.Close();
HttpStatusCode ResponseCode = ((HttpWebResponse)Response).StatusCode;
if (!ResponseCode.Equals(HttpStatusCode.OK))
{
Predict error = Newtonsoft.Json.JsonConvert.DeserializeObject <Predict>(responseLine);
return(error);
}
Predict predict = Newtonsoft.Json.JsonConvert.DeserializeObject <Predict>(responseLine);
int ind = 0;
foreach (string imgUrl in imgURLs)
{
var outputs = predict.Outputs;
outputs[ind].Data.Concepts[0].ImageName = imgUrl;
ind++;
}
return(predict);
}
catch (WebException e)
{
using (WebResponse response = e.Response)
{
HttpWebResponse httpResponse = (HttpWebResponse)response;
using (Stream data = response.GetResponseStream())
using (var reader = new StreamReader(data))
{
string text = reader.ReadToEnd();
Predict error = Newtonsoft.Json.JsonConvert.DeserializeObject <Predict>(text);
return(error);
}
}
}
return(new Predict
{
Status = new PredictStatus
{
Code = 0,
Description = "Undefined Error"
}
});
}
public static async Task <Predict> GetFolderImgsPrediction(string accessToken, string folder_path)
{
ServicePointManager.ServerCertificateValidationCallback += new RemoteCertificateValidationCallback(ValidateServerCertificate);
//
// MESSAGE CONTENT
//string postData = "client_id="+ clientId + "&client_secret="+ clientSecret + "&grant_type=client_credentials";
List <PredictInput> inputs = new List <PredictInput>();
var imgPaths = Directory.GetFiles(folder_path, "*.*", SearchOption.AllDirectories).ToList();
foreach (string imgPath in imgPaths)
{
Bitmap source = new Bitmap(imgPath);
int x = source.Width - 120;
int y = source.Height - 120;
Bitmap CroppedImage = source.Clone(new Rectangle(x, y, 120, 120), source.PixelFormat);
System.IO.MemoryStream ms = new System.IO.MemoryStream();
CroppedImage.Save(ms, System.Drawing.Imaging.ImageFormat.Jpeg);
byte[] imageArray = ms.ToArray();
//byte[] imageArray = System.IO.File.ReadAllBytes(imgPath);
string base64ImageRepresentation = Convert.ToBase64String(imageArray);
inputs.Add(new PredictInput
{
Data = new PredictImage
{
Image = new PredictImageData
{
Base64 = base64ImageRepresentation
}
}
});
}
var ins = new
{
Inputs = inputs
};
string postData = LowercaseJsonSerializer.SerializeObject(ins);
byte[] byteArray = Encoding.UTF8.GetBytes(postData);
//
// CREATE REQUEST
HttpWebRequest Request = (HttpWebRequest)WebRequest.Create(_apiEndPoint + _predictPath);
Request.Method = "POST";
Request.KeepAlive = false;
Request.ContentType = "application/json";
Request.Headers.Add("cache-control", "no-cache");
Request.Headers.Add("authorization", "Bearer " + accessToken);
Stream dataStream = await Request.GetRequestStreamAsync();
await dataStream.WriteAsync(byteArray, 0, byteArray.Length);
dataStream.Close();
//
// SEND MESSAGE
try
{
WebResponse Response = await Request.GetResponseAsync();
StreamReader Reader = new StreamReader(Response.GetResponseStream());
string responseLine = await Reader.ReadToEndAsync();
Reader.Close();
HttpStatusCode ResponseCode = ((HttpWebResponse)Response).StatusCode;
if (!ResponseCode.Equals(HttpStatusCode.OK))
{
Predict error = Newtonsoft.Json.JsonConvert.DeserializeObject <Predict>(responseLine);
return(error);
}
Predict predict = Newtonsoft.Json.JsonConvert.DeserializeObject <Predict>(responseLine);
int ind = 0;
foreach (string imgPath in imgPaths)
{
var outputs = predict.Outputs;
outputs[ind].Data.Concepts[0].ImageName = imgPath;
ind++;
}
return(predict);
}
catch (WebException e)
{
using (WebResponse response = e.Response)
{
HttpWebResponse httpResponse = (HttpWebResponse)response;
using (Stream data = response.GetResponseStream())
using (var reader = new StreamReader(data))
{
string text = reader.ReadToEnd();
Predict error = Newtonsoft.Json.JsonConvert.DeserializeObject <Predict>(text);
return(error);
}
}
}
return(new Predict
{
Status = new PredictStatus
{
Code = 0,
Description = "Undefined Error"
}
});
}
public async Task <Predict> PredictByFolderPath(string folder_path)
{
Predict response = await ClarifaiHttpClient.GetFolderImgsPrediction(AccessToken, folder_path);
return(response);
}
public async Task <Predict> PredictByImgURL(List <string> urls)
{
Predict response = await ClarifaiHttpClient.GetImgUrlPrediction(AccessToken, urls);
return(response);
}
public static ColorTypeEnum FindShapeColor(Bitmap shapeBitmap,
Predict red,
Predict green,
Predict purple)
{
List<HSL> pixels = new List<HSL>();
for (int y = 0; y < shapeBitmap.Height; y++)
{
for (int x = 0; x < shapeBitmap.Width; x++)
{
Color color = shapeBitmap.GetPixel(x, y);
pixels.Add(HSL.FromRGB(new RGB(color.R, color.G, color.B)));
}
}
HSL[] subset = pixels.OrderBy(m => m.Luminance).Take(10).ToArray();
int redCount = 0, greenCount = 0, purpleCount = 0;
foreach(HSL pixel in subset)
{
double[] input = new double[] { pixel.Hue, pixel.Saturation * 360.0, pixel.Luminance * 360.0};
double isRed = red.Compute(input);
double isPurple = purple.Compute(input);
double isGreen = green.Compute(input);
if (isRed > isPurple && isRed > isGreen)
redCount++;
else if (isPurple > isRed && isPurple > isGreen)
purpleCount++;
else if (isGreen > isRed && isGreen > isPurple)
greenCount++;
}
if (redCount > greenCount && redCount > purpleCount)
return ColorTypeEnum.Red;
else if (greenCount > redCount && greenCount > purpleCount)
return ColorTypeEnum.Green;
else if (purpleCount > redCount && purpleCount > greenCount)
return ColorTypeEnum.Purple;
else
return ColorTypeEnum.Unknown;
}
private void Start()
{
predict = gameObject.GetComponent <Predict>();
rotate = new RotateTowards();
enemyCount = new List <GameObject>();
}
// Use this for initialization
void Start()
{
canvas.SetActive (false);
normalState.SetActive (true);
drawingState.SetActive (false);
pickState.SetActive (false);
wrongState.SetActive (false);
editState.SetActive (false);
systemState.SetActive (false);
canvasOpen = false;
predictor = GameObject.FindGameObjectWithTag ("Predictor").GetComponent<Predict> ();
selector = GameObject.FindGameObjectWithTag ("ScaleSelection").GetComponent<Selection> ();
scaler = GameObject.FindGameObjectWithTag ("ScaleSelection").GetComponent<ScaleScript> ();
mapping = new int[10];
mapping [0] = 4;
mapping [1] = 6;
mapping [2] = 5;
mapping [3] = 8;
mapping [4] = 10;
mapping [5] = 1;
mapping [6] = 2;
mapping [7] = 3;
mapping [8] = 7;
mapping [9] = 9;
}
public void Predict_WhenInit_ReturnNew()
{
var output = new Predict();
Assert.IsInstanceOf(typeof(Predict), output);
}