Forecasting Principles & Practice: 2.8 Autocorrelation

Поделиться
HTML-код
  • Опубликовано: 1 дек 2024

Комментарии • 8

  • @soknya
    @soknya 2 месяца назад

    the family monster part is very helpful, many thanks!

  • @andriyosemchuk8636
    @andriyosemchuk8636 Год назад +1

    The analogy of autocorrelation with grandparents is simply brilliant.

  • @anitahaique
    @anitahaique 6 месяцев назад

    Muchas gracias profesor. Sus explicaciones son maravillosas

  • @TasiaFloissac
    @TasiaFloissac 10 месяцев назад

    This was very helpful, thank you.

  • @patrick_bateman-ty7gp
    @patrick_bateman-ty7gp 10 месяцев назад

    please correct me if im wrong :
    in the grandparents cartoon explanation for autocorrelation, at lag = 2 the reason for the correlation between the grandparent and the monster being negative is being said that they tend to be 'very different'.
    Actually negative correlation, just means they are related but in a negative sense, if one rises the other falls and vice versa, they are not very different. They are completely different if the correlation is 0.

  • @romanvasiura6705
    @romanvasiura6705 Год назад

    Can you answer for a question?
    In a case of:
    recent_production |>
    gg_lag(Beer, geom = "point", lags = 1:9) + #we can choose how many lags to show
    labs(x = "lag(Beer, k)")
    What a starting point for the lag(Beer, k)? What is a k? - Is it Q1 of 2000 or something else?
    I see we can plotted 41 lags.
    It would be amazing if you say e.g. in first lag we can see scatter plot between 2000 Q1 - to 2000 Q4... in second plot we can see - and right answer.
    I would be pleased to read correct logic of this plot.

    • @otexts
      @otexts  Год назад +2

      See the relevant section of the textbook: otexts.com/fpp3/lag-plots.html
      k is the number of lags, with values from 1, 2, .... In the first plot, k=1. In the second plot, k=2, etc.
      For example, when k=1, each point represents pairs of consecutive observations in the time series, with the value at time t on the y-axis, and the value at time t-1 on the x-axis.

    • @romanvasiura6705
      @romanvasiura6705 Год назад

      @@otexts , Thank you!
      I understood how it works.
      Maybe, for somebody like me - my conclusion also will be useful.
      I started to look from the biggest lag (lag 41) in the code I had written (lags = 38:41)
      I figured out k - is a number of lag; in the plot we point period from we substract k.
      For lag 41 it is: point Q2 x-axis: (2000 Q1) = 421; y-axis: 2010 Q2 = 374;
      For lag 40 we will have a two points:
      1 point: Q2; x-axis: (2000 Q2) = 402; y-axis: 2010 Q2 = 374;
      2-nd point: Q1 x-axis (2000 Q1) = 421; 2010 Q1 = 414;
      For lag 39... - and so on....)
      P.S. Hopefully it's right