Search code examples
rtestingnaforecastingp-value

Pvalue and Xquared is N/A in Box-Ljung test

    Box-Ljung test  
data:  <some data> 
X-squared = NA, df = 20, p-value = NA.

Pvalues and Xsquared are coming as N/A, but my data values doesn't have any zero.

Solution

  • As you don't have NA's in your data (Ljung Box test can't handle missing values), we will take a random example, assuming that it is a time series : 

    data <- c(12,45,48,31,12,7,78)

    We get (setting our lag to 20) :

    Box.test(data, lag = 20, type = "Ljung-Box", fitdf = 0)

    Box-Ljung test

data:  data
X-squared = NA, df = 20, p-value = NA

    Kinda logical to get a NA as we are setting a lag higher than the length of our time series.

    Now, we will change our lag and setting it to 1.

    Box.test(data, lag = 1, type = "Ljung-Box", fitdf = 0)

    Box-Ljung test

data:  data
X-squared = 0.30832, df = 1, p-value = 0.5787

    I think that your lag (20) is higher than the length of your time series (as it is here with the short example).

    See ?Box.test or wikipedia for more information.