I am using the Ta-Lib in C# and for the most part it works great. I am using it to calculate a number of moving averages and they calculate perfectly.
When Calling the ATR function it returns 0, always. I use the same parameters (the ones that overlap) as I do for the SMA, but still 0. Here is the code in question. Member variable Period is an integer. I have more data in the queues than the period I am asking it to calculate, and my data in the queues is clean, I've triple checked that. The return code is Success, but 0 return value.
import requests
import csv
import os
temp_file_name = 'temp_csv.csv'
url = 'http://url.to/file.csv'
download = requests.get(url)
with open(temp_file_name, 'w') as temp_file:
temp_file.writelines(download.content)
with open(temp_file_name, 'rU') as temp_file:
csv_reader = csv.reader(temp_file, dialect=csv.excel_tab)
for line in csv_reader:
print line
# delete the temp file after process
os.remove(temp_file_name)
答案 0 :(得分:1)
好的,我想了很多,终于经过了很多很多小时。在我的循环中,我检查以确保数组具有与Period相同数量的元素,如果是,我执行计算。实际上,数组需要Period + 1个元素来执行计算。这是我的整个功能。
public override void Calculate()
{
int outbegldx = 0;
int outnbelement = 0;
int marketID = 0;
TicTacTec.TA.Library.Core.RetCode retCode;
//Queue<OHLC> input = new Queue<OHLC>();
Queue<double> inputHigh = new Queue<double>();
Queue<double> inputLow = new Queue<double>();
Queue<double> inputClose = new Queue<double>();
using (SqlConnection conn = new SqlConnection(Properties.Settings.Default.DB_CONNECTION_STRING))
{
using (SqlCommand cmd = new SqlCommand("[Data].[proc_GetHistoricalData]", conn))
{
cmd.CommandType = CommandType.StoredProcedure;
cmd.Parameters.Add("@StartIndex", SqlDbType.Int);
cmd.Parameters.Add("@EndIndex", SqlDbType.Int);
cmd.Parameters.Add("@Weekly", SqlDbType.Bit);
cmd.Parameters["@StartIndex"].Value = this.startIndex;
cmd.Parameters["@EndIndex"].Value = this.endIndex;
cmd.Parameters["@Weekly"].Value = (int)this.TimeFrame;
conn.Open();
using (SqlDataReader rdr = cmd.ExecuteReader())
{
while (rdr.Read())
{
Console.WriteLine(rdr.GetInt64(0));
if (marketID != rdr.GetInt32(2))
{
marketID = rdr.GetInt32(2);
inputHigh.Clear();
inputLow.Clear();
inputClose.Clear();
}
//input.Enqueue(new OHLC(rdr.GetDecimal(4), rdr.GetDecimal(5), rdr.GetDecimal(6), rdr.GetDecimal(7)));
inputHigh.Enqueue(Convert.ToDouble(rdr.GetDecimal(5)));
inputLow.Enqueue(Convert.ToDouble(rdr.GetDecimal(6)));
inputClose.Enqueue(Convert.ToDouble(rdr.GetDecimal(7)));
//We have enough data to calculate the ATR so do so
if (inputHigh.Count >= this.Period + 1)
{
double[] outreal = new double[inputHigh.Count - 1];
int lookBack = Core.AtrLookback(this.Period);
retCode = Core.Atr(0, inputHigh.Count - 1, inputHigh.ToArray(), inputLow.ToArray(), inputClose.ToArray(), lookBack, out outbegldx, out outnbelement, outreal);
//Calculated the ATR, now write it to the database
if (retCode == Core.RetCode.Success)
{
this.Output = new AverageTrueRangeOutput(this.Period, outreal[0], outbegldx, outnbelement);
//Now insert into db
using (SqlCommand cmdInsert = new SqlCommand("[Data].[proc_InsertHistoricalIndicator]", conn))
{
cmdInsert.CommandType = CommandType.StoredProcedure;
cmdInsert.Parameters.Add("@MarketID", SqlDbType.Int);
cmdInsert.Parameters.Add("@IndicatorID", SqlDbType.Int);
cmdInsert.Parameters.Add("@Date", SqlDbType.Date);
cmdInsert.Parameters.Add("@Value", SqlDbType.Xml);
cmdInsert.Parameters.Add("@Weekly", SqlDbType.Bit);
cmdInsert.Parameters["@MarketID"].Value = rdr.GetInt32(2);
cmdInsert.Parameters["@IndicatorID"].Value = this.Id;
cmdInsert.Parameters["@Weekly"].Value = (int)this.TimeFrame;
cmdInsert.Parameters["@Date"].Value = rdr.GetDateTime(3); //IBLib.Util.GetDate(
cmdInsert.Parameters["@Value"].Value = this.Output.toXML().OuterXml;
cmdInsert.ExecuteNonQuery();
}
}
inputHigh.Dequeue();
inputLow.Dequeue();
inputClose.Dequeue();
}
}
rdr.Close();
}
}
}
}
有问题的一行是这个 - &#34; if(inputHigh.Count&gt; = this.Period + 1)&#34;而以前我只是&#34; if(inputHigh.Count&gt; = this.Period)&#34;