workday / timeseries-forecast Goto Github PK

For this dataset:
2674.8060304978917, 3371.1788109723193, 2657.161969121835, 2814.5583226655367, 3290.855749923403, 3103.622791045206, 3403.2011487950185, 2841.438925235243, 2995.312700153925, 3256.4042898633224, 2609.8702933486843, 3214.6409110870877, 2952.1736018157644, 3468.7045537306344, 3260.9227206904898, 2645.5024256492215, 3137.857549381811, 3311.3526531674556, 2929.7762119375716, 2846.05991810631, 2606.47822546165, 3174.9770937667918, 3140.910443979614, 2590.6601484185085, 3123.4299821259915, 2714.4060964141136, 3133.9561758319487, 2951.3288157912752, 2860.3114228342765, 2757.4279640677833

the next 6 data points in real life are:
3147.816496825682, 3418.2300802476093, 2856.905414401418, 3419.0312162705545, 3307.9803365878442, 3527.68377555284

Note that there is a slight trend up in the dataset. I would expect a little better result with d = 1.

predict the next 6 points:

1. use p = 2, d = 0, all others = 0, confidence = 0.8:

RMSE: 199.8163163213122
3079.5652126415816, 3018.357601612911, 2972.5923804575086, 2995.670261137454, 2998.568039880799, 2993.4644978016477

if I use p =2, d =1, all others = 0, confidence = 0.8:

RMSE: 253.6211530852703
2886.970570844559, 2835.4065710542895, 2825.3444795390224, 2862.835372496484, 2843.664918178257, 2848.710910931624

So, with d =1, it is about 20% worse measured by RMSE.

2. p =3, d = 0, confidence = 0.8:

RMSE: 215.41839087831758
3084.2187112556344, 3008.4402233252013, 2979.552888778218, 2995.6062055652237, 2996.5840132883823, 2994.1306487277516

p = 3, d = 1, confidence = 0.8:

RMSE: 290.36910432429397
2954.849845119037, 2867.9992817121565, 2862.9013638112624, 2878.2804776911416, 2896.2693454378027, 2880.143205617359

The result is about 40% worse with d = 1 measured by RMSE.

HannanRissanen.estimateARMA() calculates "yuleWalkerParams", but these params are not used anywhere in the code.
is this a "TODO" item? or maybe YuleWalker not needed (dead code)?

Hello,

We came across this Security Vulnerability: https://nvd.nist.gov/vuln/detail/CVE-2019-12134

CSV Injection (aka Excel Macro Injection or Formula Injection) exists in the export feature in Workday through 32 via a value (provided by a low-privileged user in a contact form field) that is mishandled in a CSV export.

Wanted to understand if a fix has been made on this one or if it's really a valid issue?

Thank you
Rupesh

请问，有没有知道 pdq PDQ m 这七个参数怎么求取的 ，R语言怎么求出这个曲线的pdq PDQm

I want to generate forecast for next 3 months using last 6 months data.
I tried with 9 months data and following data but getting following error:
"Failed to build ARIMA forecast: not enough data points: length=7, r=4"

Data:

int p = 3;
int d = 0;
int q = 3;
int P = 1;
int D = 1;
int Q = 0;
int m = 0;
int forecastSize = 3;
	     
final ArimaParams paramsForecast = new ArimaParams(p, d, q, P, D, Q, m);
double[] inputData = {100.0, 50.0, 100.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0};

ForecastResult forecastResult = Arima.forecast_arima(inputData, forecastSize, paramsForecast);
double[] predictedData = forecastResult.getForecast(); 

Any solution for this?

Hi!
First of all, great library, thanks for open-sourcing it!

In my project, I'd like to apply custom confidence intervals to the ARIMA predictions.
In the Readme file, you have stated that the confidence intervals can be changed in the ForecastUtil class, presumably through changing its confidence_constant_95pct constant. This, however, is not feasible if the project is used as a Maven dependency instead of using it as a local source dependency. It is, furthermore, not obvious to the layperson (or at least to me) how to calculate the constant for another confidence interval.

As far as I see it, the problem could be addressed in two ways ways:

• Provide a public, static conversion method to get the constant for a custom percentile in ForecastUtil which can then be used with the setConfInterval method of the ForecastResult class. This would maybe be the simplest solution!
• Extend the ArimaParams class with a confidence interval property, there could then maybe be another constructor with which custom percentiles can be set.

I'd be glad to contribute a solution to this if someone could give me some pointers on how to calculate the constant.

Have a great day!

I set ArimaParam as (2,2,2)(1,1,1)36
The returned value is absurdly large.
My input time series data was around 7000,while the predicted values were larger than 10,000,000.
This can't be right,can any one help me?

My understanding is that it should be get calculated on a validation set but this model forecasts values for the future (I don't provide a true 'future' values to calculated RMSE on), so how do we get that RMSE error?