Kernel ridge regression (KRR) is a technique to predict a single numeric value. KRR uses a kernel function, which compares two vectors and computes a measure of their similarity. This allows KRR to handle complex non-linear data. KRR uses the ridge technique (aka L2) to limit the size of model weights in order to prevent model overfitting.

I have a (good) sleep disorder that only allows me to sleep about 2 hours per night. This means I have a lot of time to write code. One evening, I figured I’d fine-tune my existing JavaScript implementation of KRR. I enjoy coding and I like to prevent my JavaScript skills from getting stale (and keep my brain busy). Coding in JavaScript is like solving a puzzle and is weirdly entertaining for me.

For most datasets, kernel ridge regression is one of my preferred techniques — with neural network regression a close second, and k-nearest neighbors regression used as a sanity check. KRR is complex but when it works, it often works very well.

If you have three training vectors, x0, x1, x2, the predicted value for an input vector x is y’ = (w0 * k(x, x0)) + (w1 * k(x, x1)) + (w2 * k(x, x2)). The wi are the model weights, and k() is a kernel function. Of course, you usually have a lot more than just three training items.

There are many possible kernel functions. The most common is the radial basis function (RBF). There are two main versions of RBF. The one I use is rbf(v1, v2) = exp(-1 * gamma * ||v1 – v2||^2) where gamma is a free parameter, sometimes called the inverse bandwidth, and ||v1 – v2||^2 is the squared Euclidean distance between vectors v1 and v2. If v1 = v2, the RBF value is 1.0 (maximum similarity). As the difference between v1 and v2 increases, RBF decreases towards 0.0.

Training a KRR prediction model is the process of finding the values of the weights. There is one weight per training item. There are several techniques that can be used to train a KRR model. The techniques fall into two categories: 1.) closed-form techniques that use a matrix inverse, and 2.) iterative techniques that do not use a matrix inverse. This blog describes an iterative training technique that uses plain vanilla stochastic gradient descent (SGD).

If n is the number of training items, the matrix inverse technique must compute the inverse of a n-by-n matrix, which becomes impractical at some point for large n. The SGD technique can work for arbitrarily large n, but requires two additional training parameters — a learning rate and a max_epochs — and is therefore more difficult to tune than the matrix inverse technique.

Briefly, training a KRR model using SGD looks like:

initialize weights to small random values
loop max_epochs times
  shuffle the order of the training data
  loop each training item i
    get training input x and target actual_y
    compute pred_y for x
    update associated weight so pred_y is closer to actual_y
  end-loop
  loop each weight j
    reduce weight slightly using an alpha value
  end-loop
end-loop
return curr values of weights

When using the standard matrix-inverse training technique, the “ridge” part of KRR adds a small value alpha, usually about 0.0001 or so, to the 1.0 values on the diagonal of the Kernel matrix. This helps to prevent model overfitting, when the model weights are too good and predict the training data perfectly, but then the model predicts poorly on new, previously unseen data. Adding alpha to the diagonal of the K matrix also “conditions” the matrix so that computing the matrix inverse of K is less likely to fail (all matrix inverse techniques are extremely brittle and can fail for many reasons). The reference, by researcher Max Welling, at web2.qatar.cmu.edu/~gdicaro/10315-Fall19/additional/welling-notes-on-kernel-ridge.pdf explains the math basis of this technique.

Because the SGD training technique does not use a Kernel matrix, you can’t add a small alpha to a matrix. Therefore, KRR using SGD uses weight decay to achieve the same effect. The connection between ridge/L2 regularization and weight decay is very subtle, but in a nutshell, they are functionally equivalent. See jamesmccaffrey.wordpress.com/2019/05/09/the-difference-between-neural-network-l2-regularization-and-weight-decay/.

For my demo, I used synthetic data that looks like:

-0.1660,  0.4406, -0.9998, -0.3953, -0.7065,  0.4840
 0.0776, -0.1616,  0.3704, -0.5911,  0.7562,  0.1568
-0.9452,  0.3409, -0.1654,  0.1174, -0.7192,  0.8054
. . .

The first five values on each line are predictor values. The last value is the target y value to predict. The data was generated by a 5-10-1 neural network with random weights and bias values. There are 200 training items and 40 test items.

The key calling statements in my JavaScript demo are:

. . .
  let gamma = 0.3;    // RBF param
  let alpha = 0.00001;  // decay regularization
  let krr = new KRR(gamma, alpha);

  let lrnRate = 0.05;
  let maxEpochs = 2000;
  let seed = 0;
  krr.train(trainX, trainY, lrnRate, maxEpochs, seed);
  console.log("Done ");
. . .
  let trainAcc = krr.accuracy(trainX, trainY, 0.10);
  let testAcc = krr.accuracy(testX, testY, 0.10);
. . .

The gamma parameter controls the RBF similarity kernel function, so it’s an architecture parameter. The alpha, lrnRate, and maxEpochs parameters control training. The seed parameter is used to initialize the model weights, and to shuffle the order in which the training data is processed by SGD.

The demo KRR signatures pass gamma and alpha to the constructor() and trainX, trainY, lrnRate, maxEpochs, and seed to the train() method. Because alpha is specific to training, it kind of makes more sense to pass alpha to train(), but because any form of KRR requires the alpha, it also makes sense to pass it to the constructor.

The accuracy() method scores a prediction as correct if it’s within 10% of the true target y value. The output of my demo is:

Begin Kernel Ridge Regression with SGD training

Loading train (200) and test (40) from file

First three train X:
 -0.1660   0.4406  -0.9998  -0.3953  -0.7065
  0.0776  -0.1616   0.3704  -0.5911   0.7562
 -0.9452   0.3409  -0.1654   0.1174  -0.7192

First three train y:
   0.4840
   0.1568
   0.8054

Setting RBF gamma = 0.3
Setting alpha decay = 1.0e-5

Setting SGD lrnRate = 0.050
Setting SGD maxEpochs = 2000

Creating and training KRR model using SGD
epoch =      0  MSE = 0.0129  acc = 0.2150
epoch =    400  MSE = 0.0001  acc = 0.9850
epoch =    800  MSE = 0.0001  acc = 0.9900
epoch =   1200  MSE = 0.0000  acc = 0.9850
epoch =   1600  MSE = 0.0000  acc = 0.9950
Done

Model weights:
  -1.3270   -0.7016   -0.1385   -0.6873 . . .
   0.1933    0.1438    0.4050    0.3252 . . .
. . .
  -0.2577    0.5175    0.4515   -0.0518 . . .

Computing model accuracy

Train acc (within 0.10) = 0.9900
Test acc (within 0.10) = 0.9500

Train MSE = 0.0000
Test MSE = 0.0002

Predicting for x =
  -0.1660    0.4406   -0.9998   -0.3953   -0.7065
Predicted y = 0.4942

End demo

From a practical point of view, the main challenge is to find good values for the gamma, alpha, lrnRate, and maxEpochs parameters. You can do this by manual trial and error, or by programmatic grid search.

Good fun on a sleepless night.

There are kernels and there are colonels. I’m a big fan of 1960s science fiction movies, and many of them have a military colonel character. Here are three examples.

Right: In “Space Probe Taurus” (1965), a crew of four in spaceship Hope One explore space and have several adventures, including meeting one of the strangest-looking movie aliens of all time — the alien looks like he’s sticking his tongue out like a child. Colonel Hank Stevens is the commander of the ship. By almost any standards, this is not a good movie. But I appreciate its earnestness and give it a C+ grade.

Demo program. Replace “lt” (less than), “gt”, “lte”, “gte” with Boolean operator symbols (my blog editor usually chokes on symbols).

// krr_sgd.js
// kernel ridge regression using SGD training
// node.js environment

let FS = require("fs")  // for loadTxt()

// ----------------------------------------------------------

class KRR
{
  constructor(gamma, alpha)
  {
    this.gamma = gamma;
    this.alpha = alpha;
    this.seed = 0.5;  // default init

    this.trainX;  // supplied in train()
    this.trainY;  // supplied in train()
    this.wts;     // supplied in train()
  }

  // --------------------------------------------------------

  predict(x)
  {
    let n = this.trainX.length;
    let sum = 0.0;
    for (let i = 0; i "lt" n; ++i) {
      let xx = this.trainX[i];
      let k = this.rbf(x, xx, this.gamma);
      sum += this.wts[i] * k;
    }
    return sum;
  }

  // --------------------------------------------------------

  train(trainX, trainY, lrnRate, maxEpochs, seed)
  {
    this.seed = seed + 0.5;  // quasi-random, avoid 0
    // store trainX -- needed to predict
    this.trainX = trainX;
    this.trainY = trainY;
    let n = trainX.length;  // allocate model weights
    this.wts = vecMake(n, 0.0);

    let freq = maxEpochs / 5;  // when to show progress
    let lo = -0.10; let hi = 0.10;
    for (let i = 0; i "lt" n; ++i) {
      this.wts[i] = (hi - lo) * 
        this.next() + lo;
    }

    // set up indices for shuffling
    let indices = vecMake(n, 0.0);
    for (let i = 0; i "lt" n; ++i)
      indices[i] = i;

    for (let epoch = 0; epoch "lt" maxEpochs; ++epoch) {
      this.shuffle(indices);
      for (let i = 0; i "lt" n; ++i) {
        let idx = indices[i];
        let x = trainX[idx];
        let predY = this.predict(x);
        let actualY = trainY[idx];

        // update wt assoc with x
        this.wts[idx] -= lrnRate * (predY - actualY);
      } // each item

      if (epoch % freq == 0) // show progress
      {
        let mse = this.meanSqError(trainX, trainY);
        let acc = this.accuracy(trainX, trainY, 0.10);
        let s1 = "epoch = " +
          epoch.toString().padStart(6, ' ');
        let s2 = "  MSE = " +
          mse.toFixed(4).toString();
        let s3 = "  acc = " + acc.toFixed(4).toString();
        console.log(s1 + s2 + s3);
      }

    } // each epoch

    // apply wt decay regularization
    for (let j = 0; j "lt" n; ++j)
      this.wts[j] *= (1.0 - this.alpha);

  } // train()

  // --------------------------------------------------------

  rbf(v1, v2, gamma)  // could omit gamma and use this.gamma
  {
    let n = v1.length;
    let sum = 0.0;
    for (let i = 0; i "lt" n; ++i) {
      let diff = v1[i] - v2[i];
      sum += diff * diff;
    }
    return Math.exp(-1 * gamma * sum);
  }

  // --------------------------------------------------------

  accuracy(dataX, dataY, pctClose)
  {
    let nCorrect = 0; let nWrong = 0;
    let n = dataX.length;
    
    for (let i = 0; i "lt" n; ++i) {
      let x = dataX[i];
      let actualY = dataY[i];
      let predY = this.predict(x);
      if (Math.abs(predY - actualY) "lt" 
        Math.abs(pctClose * actualY)) {
        ++nCorrect;
      }
      else {
        ++nWrong;
      }
    }
    return (nCorrect * 1.0) / (nCorrect + nWrong);
  }

  // --------------------------------------------------------

  meanSqError(dataX, dataY)
  {
    let n = dataX.length;
    let sum = 0.0;
    for (let i = 0; i "lt" n; ++i) {
      let x = dataX[i];
      let actualY = dataY[i];
      let predY = this.predict(x);
      sum += (actualY - predY) * (actualY - predY);
    }
    return sum / n;
  }

  // --------------------------------------------------------

  next()
  {
    let x = Math.sin(this.seed) * 1000;
    let result = x - Math.floor(x);  // [0.0,1.0)
    this.seed = result;  // for next call
    return result;
  }

  // --------------------------------------------------------

  nextInt(lo, hi)
  {
    let x = this.next();
    return Math.trunc((hi - lo) * x + lo);
  }

  // --------------------------------------------------------

  shuffle(indices)
  {
    // Fisher-Yates
    for (let i = 0; i "lt" indices.length; ++i) {
      let ri = this.nextInt(i, indices.length);
      let tmp = indices[ri];
      indices[ri] = indices[i];
      indices[i] = tmp;
      //indices[i] = i; // for testing
    }
  }

} // end class KRR

// ----------------------------------------------------------
// vector and matrix helper functions
// ----------------------------------------------------------

// ----------------------------------------------------------

function vecMake(n, val)
{
  let result = [];
  for (let i = 0; i "lt" n; ++i) {
    result[i] = val;
  }
  return result;
}

// ----------------------------------------------------------

function matMake(rows, cols, val)
{
  let result = [];
  for (let i = 0; i "lt" rows; ++i) {
    result[i] = [];
    for (let j = 0; j "lt" cols; ++j) {
      result[i][j] = val;
    }
  }
  return result;
}

// ----------------------------------------------------------

function vecShow(vec, dec, wid, nl)
{
  let small = 1.0 / Math.pow(10, dec);
  for (let i = 0; i "lt" vec.length; ++i) {
    let x = vec[i];
    if (Math.abs(x) "lt" small) x = 0.0  // avoid -0.00
    let xx = x.toFixed(dec);
    let s = xx.toString().padStart(wid, ' ');
    process.stdout.write(s);
    process.stdout.write(" ");
  }

  if (nl == true)
    process.stdout.write("\n");
}

// ----------------------------------------------------------

function matShow(A, dec, wid)
{
  let small = 1.0 / Math.pow(10, dec);
  let nr = A.length;
  let nc = A[0].length;
  for (let i = 0; i "lt" nr; ++i) {
    for (let j = 0; j "lt" nc; ++j) {
      let x = A[i][j];
      if (Math.abs(x) "lt" small) x = 0.0;
      let xx = x.toFixed(dec);
      let s = xx.toString().padStart(wid, ' ');
      process.stdout.write(s);
      process.stdout.write(" ");
    }
    process.stdout.write("\n");
  }
}

// ----------------------------------------------------------

function matToVec(m)
{
  let r = m.length;
  let c = m[0].length;
  let result = 	vecMake(r*c, 0.0);
  let k = 0;
  for (let i = 0; i "lt" r; ++i) {
    for (let j = 0; j "lt" c; ++j) {
      result[k++] = m[i][j];
    }
  }
  return result;
}

// ----------------------------------------------------------

function loadTxt(fn, delimit, usecols, comment) // aka matLoad
{
  let all = FS.readFileSync(fn, "utf8");  // giant string
  all = all.trim();  // strip final crlf in file
  let lines = all.split("\n");  // array of lines

  // count number non-comment lines
  let nRows = 0;
  for (let i = 0; i "lt" lines.length; ++i) {
    if (!lines[i].startsWith(comment))
      ++nRows;
  }
  let nCols = usecols.length;
  let result = matMake(nRows, nCols, 0.0); 
 
  let r = 0;  // into lines
  let i = 0;  // into result[][]
  while (r "lt" lines.length) {
    if (lines[r].startsWith(comment)) {
      ++r;  // next row
    }
    else {
      let tokens = lines[r].split(delimit);
      for (let j = 0; j "lt" nCols; ++j) {
        result[i][j] = parseFloat(tokens[usecols[j]]);
      }
      ++r;
      ++i;
    }
  }

  return result;
}

// ----------------------------------------------------------
// ----------------------------------------------------------

function main()
{
  console.log("\nBegin Kernel Ridge " +
    "Regression with SGD training ");

  // 1. load data
  console.log("\nLoading train (200) and" +
    " test (40) from file ");

  let trainFile = ".\\Data\\synthetic_train_200.txt";
  let trainX = loadTxt(trainFile, ",", [0,1,2,3,4], "#");
  let trainY = loadTxt(trainFile, ",", [5], "#");
  trainY = matToVec(trainY);
  
  let testFile = ".\\Data\\synthetic_test_40.txt";
  let testX = loadTxt(testFile, ",", [0,1,2,3,4], "#");
  let testY = loadTxt(testFile, ",", [5], "#");
  testY = matToVec(testY);

  console.log("\nFirst three train X: ");
  for (let i = 0; i "lt" 3; ++i)
    vecShow(trainX[i], 4, 8, true); // true: add newline

  console.log("\nFirst three train y: ");
  for (let i = 0; i "lt" 3; ++i)
    console.log(trainY[i].toFixed(4).toString().
    padStart(9, ' '));

  // 2. create and train KRR model
  let gamma = 0.3;    // RBF param
  let alpha = 0.00001;  // decay regularization
  console.log("\nSetting RBF gamma = " +
    gamma.toFixed(1).toString());
  console.log("Setting alpha decay = " +
    alpha.toExponential(1).toString());
  let krr = new KRR(gamma, alpha); 

  let lrnRate = 0.05;
  let maxEpochs = 2000;
  let seed = 0;
  console.log("\nSetting SGD lrnRate = " +
    lrnRate.toFixed(3).toString());
  console.log("Setting SGD maxEpochs = " +
    maxEpochs.toString());
  console.log("\nCreating and training KRR" +
    " model using SGD ");
  krr.train(trainX, trainY, lrnRate, maxEpochs, seed);
  console.log("Done ");

  // 3. show trained model weights
  console.log("\nModel weights: ");
  vecShow(krr.wts, 4, 9, true);

  // 4. evaluate model
  console.log("\nComputing model accuracy ");
  let trainAcc = krr.accuracy(trainX, trainY, 0.10);
  let testAcc = krr.accuracy(testX, testY, 0.10);

  console.log("\nTrain acc (within 0.10) = " +
    trainAcc.toFixed(4).toString());
  console.log("Test acc (within 0.10) = " +
    testAcc.toFixed(4).toString());

  let trainMSE = krr.meanSqError(trainX, trainY);
  let testMSE = krr.meanSqError(testX, testY);

  console.log("\nTrain MSE = " +
    trainMSE.toFixed(4).toString());
  console.log("Test MSE = " +
    testMSE.toFixed(4).toString());

  // 5. use model
  let x = trainX[0];
  console.log("\nPredicting for x = ");
  vecShow(x, 4, 9, true);  // add newline

  let predY = krr.predict(x);
  console.log("Predicted y = " + 
    predY.toFixed(4).toString());

  console.log("\nEnd demo");
}

main()

Training data:

# synthetic_train_200.txt
#
-0.1660,  0.4406, -0.9998, -0.3953, -0.7065,  0.4840
 0.0776, -0.1616,  0.3704, -0.5911,  0.7562,  0.1568
-0.9452,  0.3409, -0.1654,  0.1174, -0.7192,  0.8054
 0.9365, -0.3732,  0.3846,  0.7528,  0.7892,  0.1345
-0.8299, -0.9219, -0.6603,  0.7563, -0.8033,  0.7955
 0.0663,  0.3838, -0.3690,  0.3730,  0.6693,  0.3206
-0.9634,  0.5003,  0.9777,  0.4963, -0.4391,  0.7377
-0.1042,  0.8172, -0.4128, -0.4244, -0.7399,  0.4801
-0.9613,  0.3577, -0.5767, -0.4689, -0.0169,  0.6861
-0.7065,  0.1786,  0.3995, -0.7953, -0.1719,  0.5569
 0.3888, -0.1716, -0.9001,  0.0718,  0.3276,  0.2500
 0.1731,  0.8068, -0.7251, -0.7214,  0.6148,  0.3297
-0.2046, -0.6693,  0.8550, -0.3045,  0.5016,  0.2129
 0.2473,  0.5019, -0.3022, -0.4601,  0.7918,  0.2613
-0.1438,  0.9297,  0.3269,  0.2434, -0.7705,  0.5171
 0.1568, -0.1837, -0.5259,  0.8068,  0.1474,  0.3307
-0.9943,  0.2343, -0.3467,  0.0541,  0.7719,  0.5581
 0.2467, -0.9684,  0.8589,  0.3818,  0.9946,  0.1092
-0.6553, -0.7257,  0.8652,  0.3936, -0.8680,  0.7018
 0.8460,  0.4230, -0.7515, -0.9602, -0.9476,  0.1996
-0.9434, -0.5076,  0.7201,  0.0777,  0.1056,  0.5664
 0.9392,  0.1221, -0.9627,  0.6013, -0.5341,  0.1533
 0.6142, -0.2243,  0.7271,  0.4942,  0.1125,  0.1661
 0.4260,  0.1194, -0.9749, -0.8561,  0.9346,  0.2230
 0.1362, -0.5934, -0.4953,  0.4877, -0.6091,  0.3810
 0.6937, -0.5203, -0.0125,  0.2399,  0.6580,  0.1460
-0.6864, -0.9628, -0.8600, -0.0273,  0.2127,  0.5387
 0.9772,  0.1595, -0.2397,  0.1019,  0.4907,  0.1611
 0.3385, -0.4702, -0.8673, -0.2598,  0.2594,  0.2270
-0.8669, -0.4794,  0.6095, -0.6131,  0.2789,  0.4700
 0.0493,  0.8496, -0.4734, -0.8681,  0.4701,  0.3516
 0.8639, -0.9721, -0.5313,  0.2336,  0.8980,  0.1412
 0.9004,  0.1133,  0.8312,  0.2831, -0.2200,  0.1782
 0.0991,  0.8524,  0.8375, -0.2102,  0.9265,  0.2150
-0.6521, -0.7473, -0.7298,  0.0113, -0.9570,  0.7422
 0.6190, -0.3105,  0.8802,  0.1640,  0.7577,  0.1056
 0.6895,  0.8108, -0.0802,  0.0927,  0.5972,  0.2214
 0.1982, -0.9689,  0.1870, -0.1326,  0.6147,  0.1310
-0.3695,  0.7858,  0.1557, -0.6320,  0.5759,  0.3773
-0.1596,  0.3581,  0.8372, -0.9992,  0.9535,  0.2071
-0.2468,  0.9476,  0.2094,  0.6577,  0.1494,  0.4132
 0.1737,  0.5000,  0.7166,  0.5102,  0.3961,  0.2611
 0.7290, -0.3546,  0.3416, -0.0983, -0.2358,  0.1332
-0.3652,  0.2438, -0.1395,  0.9476,  0.3556,  0.4170
-0.6029, -0.1466, -0.3133,  0.5953,  0.7600,  0.4334
-0.4596, -0.4953,  0.7098,  0.0554,  0.6043,  0.2775
 0.1450,  0.4663,  0.0380,  0.5418,  0.1377,  0.2931
-0.8636, -0.2442, -0.8407,  0.9656, -0.6368,  0.7429
 0.6237,  0.7499,  0.3768,  0.1390, -0.6781,  0.2185
-0.5499,  0.1850, -0.3755,  0.8326,  0.8193,  0.4399
-0.4858, -0.7782, -0.6141, -0.0008,  0.4572,  0.4197
 0.7033, -0.1683,  0.2334, -0.5327, -0.7961,  0.1776
 0.0317, -0.0457, -0.6947,  0.2436,  0.0880,  0.3345
 0.5031, -0.5559,  0.0387,  0.5706, -0.9553,  0.3107
-0.3513,  0.7458,  0.6894,  0.0769,  0.7332,  0.3170
 0.2205,  0.5992, -0.9309,  0.5405,  0.4635,  0.3532
-0.4806, -0.4859,  0.2646, -0.3094,  0.5932,  0.3202
 0.9809, -0.3995, -0.7140,  0.8026,  0.0831,  0.1600
 0.9495,  0.2732,  0.9878,  0.0921,  0.0529,  0.1289
-0.9476, -0.6792,  0.4913, -0.9392, -0.2669,  0.5966
 0.7247,  0.3854,  0.3819, -0.6227, -0.1162,  0.1550
-0.5922, -0.5045, -0.4757,  0.5003, -0.0860,  0.5863
-0.8861,  0.0170, -0.5761,  0.5972, -0.4053,  0.7301
 0.6877, -0.2380,  0.4997,  0.0223,  0.0819,  0.1404
 0.9189,  0.6079, -0.9354,  0.4188, -0.0700,  0.1907
-0.1428, -0.7820,  0.2676,  0.6059,  0.3936,  0.2790
 0.5324, -0.3151,  0.6917, -0.1425,  0.6480,  0.1071
-0.8432, -0.9633, -0.8666, -0.0828, -0.7733,  0.7784
-0.9444,  0.5097, -0.2103,  0.4939, -0.0952,  0.6787
-0.0520,  0.6063, -0.1952,  0.8094, -0.9259,  0.4836
 0.5477, -0.7487,  0.2370, -0.9793,  0.0773,  0.1241
 0.2450,  0.8116,  0.9799,  0.4222,  0.4636,  0.2355
 0.8186, -0.1983, -0.5003, -0.6531, -0.7611,  0.1511
-0.4714,  0.6382, -0.3788,  0.9648, -0.4667,  0.5950
 0.0673, -0.3711,  0.8215, -0.2669, -0.1328,  0.2677
-0.9381,  0.4338,  0.7820, -0.9454,  0.0441,  0.5518
-0.3480,  0.7190,  0.1170,  0.3805, -0.0943,  0.4724
-0.9813,  0.1535, -0.3771,  0.0345,  0.8328,  0.5438
-0.1471, -0.5052, -0.2574,  0.8637,  0.8737,  0.3042
-0.5454, -0.3712, -0.6505,  0.2142, -0.1728,  0.5783
 0.6327, -0.6297,  0.4038, -0.5193,  0.1484,  0.1153
-0.5424,  0.3282, -0.0055,  0.0380, -0.6506,  0.6613
 0.1414,  0.9935,  0.6337,  0.1887,  0.9520,  0.2540
-0.9351, -0.8128, -0.8693, -0.0965, -0.2491,  0.7353
 0.9507, -0.6640,  0.9456,  0.5349,  0.6485,  0.1059
-0.0462, -0.9737, -0.2940, -0.0159,  0.4602,  0.2606
-0.0627, -0.0852, -0.7247, -0.9782,  0.5166,  0.2977
 0.0478,  0.5098, -0.0723, -0.7504, -0.3750,  0.3335
 0.0090,  0.3477,  0.5403, -0.7393, -0.9542,  0.4415
-0.9748,  0.3449,  0.3736, -0.1015,  0.8296,  0.4358
 0.2887, -0.9895, -0.0311,  0.7186,  0.6608,  0.2057
 0.1570, -0.4518,  0.1211,  0.3435, -0.2951,  0.3244
 0.7117, -0.6099,  0.4946, -0.4208,  0.5476,  0.1096
-0.2929, -0.5726,  0.5346, -0.3827,  0.4665,  0.2465
 0.4889, -0.5572, -0.5718, -0.6021, -0.7150,  0.2163
-0.7782,  0.3491,  0.5996, -0.8389, -0.5366,  0.6516
-0.5847,  0.8347,  0.4226,  0.1078, -0.3910,  0.6134
 0.8469,  0.4121, -0.0439, -0.7476,  0.9521,  0.1571
-0.6803, -0.5948, -0.1376, -0.1916, -0.7065,  0.7156
 0.2878,  0.5086, -0.5785,  0.2019,  0.4979,  0.2980
 0.2764,  0.1943, -0.4090,  0.4632,  0.8906,  0.2960
-0.8877,  0.6705, -0.6155, -0.2098, -0.3998,  0.7107
-0.8398,  0.8093, -0.2597,  0.0614, -0.0118,  0.6502
-0.8476,  0.0158, -0.4769, -0.2859, -0.7839,  0.7715
 0.5751, -0.7868,  0.9714, -0.6457,  0.1448,  0.1175
 0.4802, -0.7001,  0.1022, -0.5668,  0.5184,  0.1090
 0.4458, -0.6469,  0.7239, -0.9604,  0.7205,  0.0779
 0.5175,  0.4339,  0.9747, -0.4438, -0.9924,  0.2879
 0.8678,  0.7158,  0.4577,  0.0334,  0.4139,  0.1678
 0.5406,  0.5012,  0.2264, -0.1963,  0.3946,  0.2088
-0.9938,  0.5498,  0.7928, -0.5214, -0.7585,  0.7687
 0.7661,  0.0863, -0.4266, -0.7233, -0.4197,  0.1466
 0.2277, -0.3517, -0.0853, -0.1118,  0.6563,  0.1767
 0.3499, -0.5570, -0.0655, -0.3705,  0.2537,  0.1632
 0.7547, -0.1046,  0.5689, -0.0861,  0.3125,  0.1257
 0.8186,  0.2110,  0.5335,  0.0094, -0.0039,  0.1391
 0.6858, -0.8644,  0.1465,  0.8855,  0.0357,  0.1845
-0.4967,  0.4015,  0.0805,  0.8977,  0.2487,  0.4663
 0.6760, -0.9841,  0.9787, -0.8446, -0.3557,  0.1509
-0.1203, -0.4885,  0.6054, -0.0443, -0.7313,  0.4854
 0.8557,  0.7919, -0.0169,  0.7134, -0.1628,  0.2002
 0.0115, -0.6209,  0.9300, -0.4116, -0.7931,  0.4052
-0.7114, -0.9718,  0.4319,  0.1290,  0.5892,  0.3661
 0.3915,  0.5557, -0.1870,  0.2955, -0.6404,  0.2954
-0.3564, -0.6548, -0.1827, -0.5172, -0.1862,  0.4622
 0.2392, -0.4959,  0.5857, -0.1341, -0.2850,  0.2470
-0.3394,  0.3947, -0.4627,  0.6166, -0.4094,  0.5325
 0.7107,  0.7768, -0.6312,  0.1707,  0.7964,  0.2757
-0.1078,  0.8437, -0.4420,  0.2177,  0.3649,  0.4028
-0.3139,  0.5595, -0.6505, -0.3161, -0.7108,  0.5546
 0.4335,  0.3986,  0.3770, -0.4932,  0.3847,  0.1810
-0.2562, -0.2894, -0.8847,  0.2633,  0.4146,  0.4036
 0.2272,  0.2966, -0.6601, -0.7011,  0.0284,  0.2778
-0.0743, -0.1421, -0.0054, -0.6770, -0.3151,  0.3597
-0.4762,  0.6891,  0.6007, -0.1467,  0.2140,  0.4266
-0.4061,  0.7193,  0.3432,  0.2669, -0.7505,  0.6147
-0.0588,  0.9731,  0.8966,  0.2902, -0.6966,  0.4955
-0.0627, -0.1439,  0.1985,  0.6999,  0.5022,  0.3077
 0.1587,  0.8494, -0.8705,  0.9827, -0.8940,  0.4263
-0.7850,  0.2473, -0.9040, -0.4308, -0.8779,  0.7199
 0.4070,  0.3369, -0.2428, -0.6236,  0.4940,  0.2215
-0.0242,  0.0513, -0.9430,  0.2885, -0.2987,  0.3947
-0.5416, -0.1322, -0.2351, -0.0604,  0.9590,  0.3683
 0.1055,  0.7783, -0.2901, -0.5090,  0.8220,  0.2984
-0.9129,  0.9015,  0.1128, -0.2473,  0.9901,  0.4776
-0.9378,  0.1424, -0.6391,  0.2619,  0.9618,  0.5368
 0.7498, -0.0963,  0.4169,  0.5549, -0.0103,  0.1614
-0.2612, -0.7156,  0.4538, -0.0460, -0.1022,  0.3717
 0.7720,  0.0552, -0.1818, -0.4622, -0.8560,  0.1685
-0.4177,  0.0070,  0.9319, -0.7812,  0.3461,  0.3052
-0.0001,  0.5542, -0.7128, -0.8336, -0.2016,  0.3803
 0.5356, -0.4194, -0.5662, -0.9666, -0.2027,  0.1776
-0.2378,  0.3187, -0.8582, -0.6948, -0.9668,  0.5474
-0.1947, -0.3579,  0.1158,  0.9869,  0.6690,  0.2992
 0.3992,  0.8365, -0.9205, -0.8593, -0.0520,  0.3154
-0.0209,  0.0793,  0.7905, -0.1067,  0.7541,  0.1864
-0.4928, -0.4524, -0.3433,  0.0951, -0.5597,  0.6261
-0.8118,  0.7404, -0.5263, -0.2280,  0.1431,  0.6349
 0.0516, -0.8480,  0.7483,  0.9023,  0.6250,  0.1959
-0.3212,  0.1093,  0.9488, -0.3766,  0.3376,  0.2735
-0.3481,  0.5490, -0.3484,  0.7797,  0.5034,  0.4379
-0.5785, -0.9170, -0.3563, -0.9258,  0.3877,  0.4121
 0.3407, -0.1391,  0.5356,  0.0720, -0.9203,  0.3458
-0.3287, -0.8954,  0.2102,  0.0241,  0.2349,  0.3247
-0.1353,  0.6954, -0.0919, -0.9692,  0.7461,  0.3338
 0.9036, -0.8982, -0.5299, -0.8733, -0.1567,  0.1187
 0.7277, -0.8368, -0.0538, -0.7489,  0.5458,  0.0830
 0.9049,  0.8878,  0.2279,  0.9470, -0.3103,  0.2194
 0.7957, -0.1308, -0.5284,  0.8817,  0.3684,  0.2172
 0.4647, -0.4931,  0.2010,  0.6292, -0.8918,  0.3371
-0.7390,  0.6849,  0.2367,  0.0626, -0.5034,  0.7039
-0.1567, -0.8711,  0.7940, -0.5932,  0.6525,  0.1710
 0.7635, -0.0265,  0.1969,  0.0545,  0.2496,  0.1445
 0.7675,  0.1354, -0.7698, -0.5460,  0.1920,  0.1728
-0.5211, -0.7372, -0.6763,  0.6897,  0.2044,  0.5217
 0.1913,  0.1980,  0.2314, -0.8816,  0.5006,  0.1998
 0.8964,  0.0694, -0.6149,  0.5059, -0.9854,  0.1825
 0.1767,  0.7104,  0.2093,  0.6452,  0.7590,  0.2832
-0.3580, -0.7541,  0.4426, -0.1193, -0.7465,  0.5657
-0.5996,  0.5766, -0.9758, -0.3933, -0.9572,  0.6800
 0.9950,  0.1641, -0.4132,  0.8579,  0.0142,  0.2003
-0.4717, -0.3894, -0.2567, -0.5111,  0.1691,  0.4266
 0.3917, -0.8561,  0.9422,  0.5061,  0.6123,  0.1212
-0.0366, -0.1087,  0.3449, -0.1025,  0.4086,  0.2475
 0.3633,  0.3943,  0.2372, -0.6980,  0.5216,  0.1925
-0.5325, -0.6466, -0.2178, -0.3589,  0.6310,  0.3568
 0.2271,  0.5200, -0.1447, -0.8011, -0.7699,  0.3128
 0.6415,  0.1993,  0.3777, -0.0178, -0.8237,  0.2181
-0.5298, -0.0768, -0.6028, -0.9490,  0.4588,  0.4356
 0.6870, -0.1431,  0.7294,  0.3141,  0.1621,  0.1632
-0.5985,  0.0591,  0.7889, -0.3900,  0.7419,  0.2945
 0.3661,  0.7984, -0.8486,  0.7572, -0.6183,  0.3449
 0.6995,  0.3342, -0.3113, -0.6972,  0.2707,  0.1712
 0.2565,  0.9126,  0.1798, -0.6043, -0.1413,  0.2893
-0.3265,  0.9839, -0.2395,  0.9854,  0.0376,  0.4770
 0.2690, -0.1722,  0.9818,  0.8599, -0.7015,  0.3954
-0.2102, -0.0768,  0.1219,  0.5607, -0.0256,  0.3949
 0.8216, -0.9555,  0.6422, -0.6231,  0.3715,  0.0801
-0.2896,  0.9484, -0.7545, -0.6249,  0.7789,  0.4370
-0.9985, -0.5448, -0.7092, -0.5931,  0.7926,  0.5402

Test data:

# synthetic_test_40.txt
#
 0.7462,  0.4006, -0.0590,  0.6543, -0.0083,  0.1935
 0.8495, -0.2260, -0.0142, -0.4911,  0.7699,  0.1078
-0.2335, -0.4049,  0.4352, -0.6183, -0.7636,  0.5088
 0.1810, -0.5142,  0.2465,  0.2767, -0.3449,  0.3136
-0.8650,  0.7611, -0.0801,  0.5277, -0.4922,  0.7140
-0.2358, -0.7466, -0.5115, -0.8413, -0.3943,  0.4533
 0.4834,  0.2300,  0.3448, -0.9832,  0.3568,  0.1360
-0.6502, -0.6300,  0.6885,  0.9652,  0.8275,  0.3046
-0.3053,  0.5604,  0.0929,  0.6329, -0.0325,  0.4756
-0.7995,  0.0740, -0.2680,  0.2086,  0.9176,  0.4565
-0.2144, -0.2141,  0.5813,  0.2902, -0.2122,  0.4119
-0.7278, -0.0987, -0.3312, -0.5641,  0.8515,  0.4438
 0.3793,  0.1976,  0.4933,  0.0839,  0.4011,  0.1905
-0.8568,  0.9573, -0.5272,  0.3212, -0.8207,  0.7415
-0.5785,  0.0056, -0.7901, -0.2223,  0.0760,  0.5551
 0.0735, -0.2188,  0.3925,  0.3570,  0.3746,  0.2191
 0.1230, -0.2838,  0.2262,  0.8715,  0.1938,  0.2878
 0.4792, -0.9248,  0.5295,  0.0366, -0.9894,  0.3149
-0.4456,  0.0697,  0.5359, -0.8938,  0.0981,  0.3879
 0.8629, -0.8505, -0.4464,  0.8385,  0.5300,  0.1769
 0.1995,  0.6659,  0.7921,  0.9454,  0.9970,  0.2330
-0.0249, -0.3066, -0.2927, -0.4923,  0.8220,  0.2437
 0.4513, -0.9481, -0.0770, -0.4374, -0.9421,  0.2879
-0.3405,  0.5931, -0.3507, -0.3842,  0.8562,  0.3987
 0.9538,  0.0471,  0.9039,  0.7760,  0.0361,  0.1706
-0.0887,  0.2104,  0.9808,  0.5478, -0.3314,  0.4128
-0.8220, -0.6302,  0.0537, -0.1658,  0.6013,  0.4306
-0.4123, -0.2880,  0.9074, -0.0461, -0.4435,  0.5144
 0.0060,  0.2867, -0.7775,  0.5161,  0.7039,  0.3599
-0.7968, -0.5484,  0.9426, -0.4308,  0.8148,  0.2979
 0.7811,  0.8450, -0.6877,  0.7594,  0.2640,  0.2362
-0.6802, -0.1113, -0.8325, -0.6694, -0.6056,  0.6544
 0.3821,  0.1476,  0.7466, -0.5107,  0.2592,  0.1648
 0.7265,  0.9683, -0.9803, -0.4943, -0.5523,  0.2454
-0.9049, -0.9797, -0.0196, -0.9090, -0.4433,  0.6447
-0.4607,  0.1811, -0.2389,  0.4050, -0.0078,  0.5229
 0.2664, -0.2932, -0.4259, -0.7336,  0.8742,  0.1834
-0.4507,  0.1029, -0.6294, -0.1158, -0.6294,  0.6081
 0.8948, -0.0124,  0.9278,  0.2899, -0.0314,  0.1534
-0.1323, -0.8813, -0.0146, -0.0697,  0.6135,  0.2386