Auto ARIMA in R

Поділитися
Вставка
  • Опубліковано 20 січ 2025

КОМЕНТАРІ • 22

  • @eshwarphani579
    @eshwarphani579 4 роки тому

    That was quick and best . you are the best mate . thanks a lot for this video.

  • @nelsonmosioma5305
    @nelsonmosioma5305 Рік тому

    A great explanation. Thank you.

  • @linuodeye8232
    @linuodeye8232 3 роки тому

    Thank You very Much! I am new in using r and your video is very helpful for my project!

  • @chrisogbekhiulu9923
    @chrisogbekhiulu9923 3 роки тому +2

    Great videos and thanks for this. Could I check why moved on to forecasting when the p-value is < 0.05? Isn't that requirement key?

  • @maryjoydenopol4158
    @maryjoydenopol4158 2 роки тому

    Thanks!

  • @joeo5533
    @joeo5533 2 роки тому +1

    Hello, I'm assuming we interpret "s.e." to mean standard error, and for data sets that aren't small, we could calculate our 95% confidence intervals for our coefficients via s.e.*1.96. Is this correct?

    • @joeo5533
      @joeo5533 2 роки тому

      So this approximately right, but we can get R to calculate it by dropping our model into the confint() funtion. Like:
      y

  • @mssankar3968
    @mssankar3968 3 роки тому +1

    Thank you! very helpful video! While I was trying to use the auto.arima function with seasonality on my data (fitarima

  • @nidhidhankhar8220
    @nidhidhankhar8220 Рік тому

    What's the code for formulating the model of arima with drift
    Include.drift=true is not working.
    Is there any pre package we have to install to perform drift code

  • @footbaht6401
    @footbaht6401 3 роки тому

    Thank you for your valuable tutorial! I just wanna ask if adding some dummy variables to account for those years with high inflation on top of the ARIMA model is a sensible choice to improve the model.

  • @leandanielvillareal3352
    @leandanielvillareal3352 2 роки тому

    My problem with auto.arima() function is that sometimes, it chooses a model that residuals are still not white noise. Do you have an idea how it can be prevented? Thank you so much for this tutorial.

  • @tansutazegul8297
    @tansutazegul8297 Рік тому

    Why Didnt you check whether the data is stationary?

  • @parasnathverma1405
    @parasnathverma1405 3 роки тому

    Sir, while forecasting my data set, it shows point forecast as 0 for all the next year. How to solve this issue. Please guide.

  • @nicoletabon7740
    @nicoletabon7740 Рік тому

    does SARMA has a different code with ARMA or SARIMA? TIA!

  • @allanmsiska2163
    @allanmsiska2163 3 роки тому

    Really helpful mans. Thank you

  • @allanmsiska2163
    @allanmsiska2163 3 роки тому

    Say am trying to forecast alphas and betas for stocks, will the procedure be the same?

  • @cesiacotache9961
    @cesiacotache9961 4 роки тому +1

    Thank you! I really like your videos. May I ask you how could I manage 2 variables with auto.arima? Let's say I want to predict surgical deaths under the influence of COVID-19 incidence. I would like to make predictions for both in one graph, but "COVID-19 incidence" is my predictor variable and "surgical deaths" is my response variable. Thank you!

    • @JustinEloriaga
      @JustinEloriaga  4 роки тому +3

      Hi, thanks for your comment. To my knowledge, auto.arima() only accepts one variable to forecast at a time. If you would want to predict both simultaneously or with a certain response relationship, you might want to consider a multivariate forecasting model such as the VAR. Alternatively, if you are just after the forecast, I believe you can run auto.arima() on each variable separately and then just graph the forecast using some cbind() grouping. Hope this helps.

    • @cesiacotache9961
      @cesiacotache9961 4 роки тому +1

      @@JustinEloriaga thank you! I will try as you recommended.

    • @chrisogbekhiulu9923
      @chrisogbekhiulu9923 3 роки тому

      @@cesiacotache9961 This will be ARIMAX which allows you to add an external regressor. So you add your regressor in the "xreg" argument in arima () or auto.arima () functions in R.

  • @bobbybaylonjr7403
    @bobbybaylonjr7403 4 роки тому

    thank you. Hope you could share the file and the code. I tried accessing it, but it returns an error message