Calculated-scale-5-second-data-set-calculates-to-1-tick-Solved-Data-Error

2024-10-27T18:53:00Z - 2024-10-27T18:53:00Z ago

Calculated scale: 5 second data set calculates to 1 tick [Solved - Data Error]

Hi,
The data shown below is mostly spaced at 5 seconds. And sometimes 10 seconds.
This is Y! data for Oct 14, 2024. There are supposed to be 12 x 390 = 4,680 data points.
But there are only 3,535. So there are many 10 second spaces/
The calculated scale is 1 tick.
Question 1) How can I force it to be 5 seconds?
Question 2) What is meant by 1 tick?
When I set the Strategy Settings to 1 tick, it correctly finds this data set.
So I have a solution but here is the problem with this:
I will be running a 1 min strategy along with a 5 sec strategy.
If both data sets are calculated to be 1 tick then how will strategies find their own 1 tick data sets?
Thanks!

Solved

3 Replies

Bookmark

Notify

Sort

haytac8

2024-10-27T18:54:32Z - 2024-10-27T18:54:32Z ago

I must add this 1 tick for both 1 min and 5 sec is not a common occurrence.
So far, all 1 minute data sets are correctly calculated to be 1 min scale.

Glitch8

( 11.74% )

2024-10-27T19:32:37Z - 2024-10-27T19:32:37Z ago

We’d have to take a look at the data file, could you email it to support@wealth-lab.com?

Cone8

( 6.71% )

2024-10-28T17:02:32Z - 2024-10-28T17:02:32Z ago

Your data file included the following records with the same DateTime(s) as the record that preceeded them., so we have no choice than to call it a Tick source.

CODE:
^SPX   10/14/2024   11:07:11   5846.31   0
^SPX   10/14/2024   11:07:11   5846.31   0
^SPX   10/14/2024   11:07:11   5846.31   0
^SPX   10/14/2024   11:07:11   5846.31   0
^SPX   10/14/2024   11:07:11   5846.31   0
^SPX   10/14/2024   14:06:51   5856.25   0
^SPX   10/14/2024   14:06:56   5856.26   0
^SPX   10/14/2024   14:08:46   5859.72   0
^SPX   10/14/2024   14:08:51   5859.2   0
^SPX   10/14/2024   14:09:11   5858.29   0
^SPX   10/14/2024   14:11:21   5857.56   0
^SPX   10/14/2024   14:11:36   5857.55   0
^SPX   10/14/2024   14:14:06   5858.4   0

You can use this code (which gave the output above), to filter out records that are less than 4 seconds from the previous record. By default it looks for the data file in the WL Data folder, and writes new "filtered" corrected file there.

CODE:
using WealthLab.Backtest;
using System;
using WealthLab.Core;
using WealthLab.Indicators;
using System.Collections.Generic;
using System.IO;

namespace WealthScript4 
{
    public class RemoveDuplicateDates : UserStrategyBase
    {
        public override void Initialize(BarHistory bars)
        {      
        }

      public override void BacktestBegin()
      {
         string symbol = "^SPX";
         string fmt = "M/d/yyyy HH:mm:ss";
         List<string> filtered = [];
         
         DateTime dte = DateTime.MinValue;
         string[] lines = File.ReadAllLines(Path.Combine(WLHost.Instance.DataFolder, $"{symbol}.txt"));

         DateTime lastDte = DateTime.MinValue; 
         foreach (string line in lines)
         {
            if (line.Contains("Date") || line.Contains("Close"))      // skip a header
               continue;

            string[] tokens = line.Split('\t');
            if (tokens.Length < 3)
               continue;

            string tokendte = $"{tokens[1]} {tokens[2]}";
            
            // if more than 4 seconds from the last date, add it to the filtered list, otherwise ignore this line
            if (DateTime.TryParse(tokendte, out dte))
            {
               if ((dte - lastDte).TotalSeconds > 4)
               {
                  filtered.Add(line);
                  lastDte = dte;
               }
               else
               {
                  // records with a DateTime less than 4 seconds from the previous record
                  WriteToDebugLog($"{line}");
               }
            }
         }

         // write the filtered list to a new file
         File.WriteAllLines(Path.Combine(WLHost.Instance.DataFolder, $"{symbol}-filtered.txt"), filtered);
      }

        public override void Execute(BarHistory bars, int idx)
        { }      
    }
}

Best Answer

Bookmark

Notify

Sort