在 C# 中使用 Deedle 进行成对匹配和开窗

Question

假设我们要计算一些水果的价格如何变化。从 CSV 文件开始：

Day,Name,Kind,Price
2019-09-04,"apple","red delicious",63.09
2019-09-04,"apple","ginger crisp",52.14
2019-09-04,"orange","navel",41.18
2019-09-03,"apple","red delicious",63.07
2019-09-03,"apple","ginger crisp",52.11
2019-09-03,"orange","navel",41.13
2019-09-02,"apple","red delicious",63.00
2019-09-02,"apple","ginger crisp",52.00
2019-09-02,"orange","navel",41.00

对于未知数量的水果和品种，我们可以读取数据框并构建一个额外的列用于匹配。

var fruits_file = Path.Combine(root, "fruits.csv");
Deedle.Frame<int, string> df = Frame.ReadCsv(fruits_file);

Series<int, string> name = df.GetColumn<string>("Name");
Series<int, string> kind = df.GetColumn<string>("Kind");

var namekind = name.ZipInner(kind).Select(t => t.Value.Item1 + t.Value.Item2);
df.AddColumn("NameKind", namekind);

但问题依旧。 Deedle.Series.Window() 和 Deedle.Series.Pairwise() 可以执行一阶差分，但不能基于某些字符串（同名类）进行匹配。

复制列 LastPrice 并随后计算 Change 的正确方法是什么？

Day,Name,Kind,Price,LastPrice,Change
2019-09-04,"apple","red delicious",63.09,63.07,0.02
2019-09-04,"apple","ginger crisp",52.14,52.11,0.03
2019-09-04,"orange","navel",41.18,41.13,0.05
2019-09-03,"apple","red delicious",63.07,63.00,0.07
2019-09-03,"apple","ginger crisp",52.11,52.00,0.11
2019-09-03,"orange","navel",41.13,41.00,0.13
2019-09-02,"apple","red delicious",63.00,,
2019-09-02,"apple","ginger crisp",52.00,,
2019-09-02,"orange","navel",41.00,,

Answer 1

查看下面的代码：

using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;
using System.Data;
using System.IO;

namespace ConsoleApplication137
{
    class Program
    {
        const string FILENAME = @"c:\temp\test.csv";
        static void Main(string[] args)
        {
            Fruit fruit = new Fruit(FILENAME);
            Fruit.PrintFruits();
            Console.ReadLine();
        }



    }
    public class Fruit
    {
        public static List<Fruit> fruits { get; set; }
        public DateTime day { get; set; }
        public string name { get; set; }
        public string kind { get; set; }
        public decimal price { get; set; }

        public Fruit() { }
        public Fruit(string filename)
        {
            int count = 0;
            StreamReader reader = new StreamReader(filename);
            string line = "";

            while ((line = reader.ReadLine()) != null)
            {
                if (++count > 1)
                {
                    string[] splitLine = line.Split(new char[] { ',' }).ToArray();

                    Fruit newFruit = new Fruit();
                    if (fruits == null) fruits = new List<Fruit>();
                    fruits.Add(newFruit);

                    newFruit.day = DateTime.Parse(splitLine[0]);
                    newFruit.name = splitLine[1];
                    newFruit.kind = splitLine[2];
                    newFruit.price  = decimal.Parse(splitLine[3]);

                }
            }
            reader.Close();

        }
        public static void PrintFruits()
        {
            var groups = fruits.OrderBy(x => x.day)
                .GroupBy(x => new { name = x.name, kind = x.kind })
                .ToList();

            foreach (var group in groups)
            {
                for (int i = 0; i < group.Count() - 1; i++)
                {
                    Console.WriteLine("Old Date : '{0}', New Date : '{1}', Name : '{2}', Kind : '{3}', Old Price '{4}', New Price '{5}', Delta Price '{6}'",
                        group.ToList()[i].day.ToString("yyyy-MM-dd"),
                        group.ToList()[i + 1].day.ToString("yyyy-MM-dd"),
                        group.ToList()[i].name,
                        group.ToList()[i].kind,
                        group.ToList()[i].price.ToString(),
                        group.ToList()[i + 1].price.ToString(),
                        (group.ToList()[i + 1].price - group.ToList()[i].price).ToString()
                        );


                }
            }
        }

    }
}

在 C# 中使用 Deedle 进行成对匹配和开窗

Pairwise matching and windowing using Deedle in C#

c#

dataframe

deedle