ml_preprocessing 1.1.0

Implementaion of popular algorithms of data preprocessing for machine learning

example/main.dart

import 'dart:async';

import 'package:ml_preprocessing/categorical_data_encoder_type.dart';
import 'package:ml_preprocessing/float32x4_csv_ml_data.dart';
import 'package:tuple/tuple.dart';

Future main() async {
  // Let's create data container from the csv file,
  // `labelIdx: 3` means that the label (dependent variable in terms of Machine Learning) column of the dataset is its
  // third column
  // `headerExists: true` means, that our csv-file has a header row
  // `categoryNameToEncoder: {...}` means, that we want to encode values of `position`-column with one-hot encoder
  // and column `country` will be encoded with Ordinal encoder
  // `rows: [Tuple2<int, int>(0, 6)]` means, that we want to read range of the csv's rows from 0 to 6th line
  // `columns: [Tuple2<int, int>(0, 3)]` means, that we want to read range of the csv's columns from 0 to third columns
  final data = Float32x4CsvMLData.fromFile('example/dataset.csv', labelIdx: 3,
    headerExists: true,
    categoryNameToEncoder: {
      'position': CategoricalDataEncoderType.oneHot,
      'country': CategoricalDataEncoderType.ordinal,
    },
    rows: [Tuple2<int, int>(0, 6)],
    columns: [Tuple2<int, int>(0, 3)],
  );

  // Let's read the header of the dataset, preprocessed features and labels
  final header = await data.header;
  final features = await data.features;
  final labels = await data.labels;

  // And print the result
  print(header);
  print(features);
  print(labels);

  // That's, actually, all you have to do to use data further in different applications (e.g., in Machine Learning)
}