Question about ARIMA forecast

Hi, I am a newer of SuanShuers and I want to use ARIMA model in SuanShu to forecast next value of time series just like last topic “ARIMA forecast”. After reading the reply of last topic, I have a try using my own data and the codes are listed below. However, some new questions appear.

My Data:

1.5959709882736206 <br />
0.7300914525985718 <br />
2.0011744499206543 <br />
3.6755871772766113 <br />
0.8066112399101257 <br />
1.3413848876953125 <br />
3.371157646179199 <br />
0.4400146007537842 <br />
2.637667655944824 <br />
2.1453769207000732 <br />
2.341433048248291 <br />
2.3429665565490723 <br />
1.1187453269958496 <br />
1.4169363975524902 <br />
3.328829050064087 <br />
4.157748699188232 <br />
3.9255290031433105 <br />
2.7843635082244873

Source Code:

try{<br />
//read time series from file to ArrayList<br />
BufferedReader br = new BufferedReader(new FileReader("arimademo.txt"));<br />
String line = null;<br />
ArrayList<double> arr = new ArrayList</double><double>();<br />
while((line = br.readLine()) != null){<br />
arr.add(Double.valueOf(line));<br />
}<br />
//get the last value<br />
double last = arr.get(arr.size()-1);<br />
//remove last value<br />
arr.remove(arr.size()-1);<br />
//new a double array<br />
double[] series = new double[arr.size()];<br />
for(int i = 0; i < arr.size(); i++){<br />
series<i> = arr.get(i);<br />
}<br />
//build ARIMA model and then forecast next value<br />
ConditionalSumOfSquares instance = new ConditionalSumOfSquares(series, 1, 1, 1);<br />
ARIMAModel arima = instance.getModel();<br />
ARIMAForecastMultiStep forecast = new ARIMAForecastMultiStep(new SimpleTimeSeries(series), arima, 1);<br />
System.out.println("prediction = "+forecast.xHat());<br />
System.out.println("error = "+forecast.var());<br />
System.out.println("real = "+last);<br />
br.close();<br />
}catch(Exception e){<br />
e.printStackTrace();<br />
}</i></double>

and the output is:

prediction = 4.246483542430577<br />
error = 0.603935593121584<br />
real = 2.7843635082244873

My questions are:

1. the forecast accuracy is too low, prediction value is 4.246 but real value is 2.78
2. how to determine p,d,q automatically?
3. why error != |real-prediction|?

Thanks for your kindly attention and Waitng for your help.