Automated, fast, minimalist Curve-Fitting tool for the browser and node.js.
Check the demo site built with curfi.js.
Using npm:
$ npm install curfiUsing CDN:
<script src="https://unpkg.com/curfi@1.0.3/curfi.min.js"></script>const curfi = require("curfi");
// create an instance of curfi
let model = new curfi();
// dataset
// xi, yi
// 1, 0.5
// 2, 2.5
// 3, 2.0
// 4, 4.0
// 5, 3.5
// 6, 6.0
// 7, 5.5
let trainX = [[1, 2, 3, 4, 5, 6, 7]];
let trainY = [[0.5, 2.5, 2.0, 4.0, 3.5, 6.0, 5.5]];
// auto fit the curve for the dataset
model.AutoTrain(trainX, trainY, null, null, 3);
// model.predict() for prediction
let prediction = model.predict([[8]]);
console.log(prediction); // [[ 6.428571428569853 ]]<!DOCTYPE html>
<html>
<head></head>
<body>
<script src="https://unpkg.com/curfi/curfi.min.js"></script>
<script>
// create an instance of curfi
let model = new curfi();
// dataset
let trainX = [[1, 2, 3, 4, 5, 6, 7]];
let trainY = [[0.5, 2.5, 2.0, 4.0, 3.5, 6.0, 5.5]];
let testX = [[1, 8]];
let testY = [[0.762, 7.265]];
// auto fit the curve for the dataset
model.AutoTrain(trainX, trainY, testX, testY, 3);
// model.predict() for prediction
console.log(model.predict([[8]])); // // [[ 6.428571428569853 ]]
</script>
</body>
</html>To check how good a model is:
model.r2_score(y_true, y_pred);To save a trained model for reuse
model.saveModel(); // creates a download in browserSave model with a model name
model.saveModel("myModel"); // will download myModel.jsonLoad a saved model
model.loadModel("myModel.json"); // file reader api to open the .json filelet cf = new curfi();
cf.modelEqnnHTML(); // returns HTML of the curve equation
cf.round(num, digits); // Round a number up to digits after decimal
cf.round3(num); // Round a number up to 3 digits after decimal
cf.round2(num); // Round a number up to 2 digits after decimal
cf.matrix_multiply(a, b); // returns Matrix Multiply of a and b
cf.matrix_transpose(a); // returns Matrix Transpose of a
cf.matrix_invert(a); // returns Matrix Inverse of aPull requests are welcome. For major changes, please open an issue first to discuss what you would like to change.
Please make sure to update tests as appropriate.