Why does R segmented package's predict function return an error?-CodePudding

I am using the segmented package to create a linear model with breakpoints but when I use the predict function it returns the following error:

Error in matrix(x.values, nrow = n, ncol = k, byrow = FALSE) : 
 'data' must be of a vector type, was 'NULL'

Here is my data:

 > dput(spring.plot)
structure(list(Year = c(1995, 1996, 1997, 1998, 1999, 2000, 2001, 
2002, 2003, 2004, 2005, 2006, 2007, 2009, 2010, 2011, 2012, 2014, 
2017, 2018), cor.d13C.x = c(-20.1731091673535, -19.9205871049139, 
-20.6374520869892, -20.1851878984857, -20.4450023031073, -20.7036033075104, 
-21.0707619356372, -20.7402193699317, -21.54775822157, -21.1569885512026, 
-21.3501796817414, -21.6191130116038, -21.6492696573744, -21.6213435694402, 
-21.5670838470492, -21.6346790950157, -22.1278529495184, -22.0151931682557, 
-22.1203815975813, -21.8421435334785), cor.d13C.y = c(0.187648385511187, 
0.303333143562945, 0.322078617191699, 0.216439794978181, 0.183331753701332, 
0.444176460919643, 0.287401841353699, 0.268731238031834, 0.18822947285027, 
0.133288158421617, 0.264489364220891, 0.443636742676881, 0.228538121333274, 
0.326033302668887, 0.396003654535805, 0.272766639830925, 0.549447170011753, 
0.245148399918538, 0.378003453281401, 0.31552762151098), d15N.x = c(9.2105, 
9.636, 9.7571, 10.1341111111111, 9.8055, 10.06425, 9.8555, 10.4192, 
9.8879, 10.5179, 10.545625, 10.3765, 10.5826, 10.3744444444444, 
10.2484, 9.5749, 9.52, 9.873, 10.0519, 10.8419), d15N.y = c(0.357452172402904, 
0.418821096783712, 0.286839269742946, 0.367195262920304, 0.312860725137134, 
0.244967490554505, 0.324268150352965, 0.459335703912606, 0.502120049833149, 
0.264608914773818, 0.202960895248321, 0.499248379511085, 0.590404607028096, 
0.471943882021769, 0.448001289680683, 0.278745383618974, 0.552719941139573, 
0.484833734387367, 0.547101341414712, 0.252805348923721), PerN.x = c(12.9007502555847, 
12.8167667388916, 12.6170592308044, 12.6319534513685, 13.5568219184875, 
12.5970660448074, 12.5292349815369, 12.3908371925354, 12.8273427963257, 
12.3977887153625, 13.0017735958099, 12.3305891036987, 12.5119565963745, 
13.4092265235053, 12.9288623809814, 13.1139984130859, 12.9217474937439, 
12.8078309165107, 13.5686268806458, 13.0825483322144), PerN.y = c(0.561407252949615, 
0.589777982603924, 0.403203366840876, 0.696942305294126, 0.34608685502915, 
0.455015723273848, 0.636465123996367, 0.556050554528199, 0.469391267667275, 
0.343232155115455, 0.45609833258894, 0.430321356385325, 0.540299777960146, 
0.528607495036965, 0.603123576500966, 0.549651250473174, 0.384079375553638, 
0.51738099428862, 0.365028403333915, 0.361264153892095), PerC.x = c(46.6922031402588, 
45.7764064788818, 46.4129936218262, 46.5377320183648, 44.9868389129639, 
43.2995181083679, 45.2353702545166, 45.0508312225342, 45.2698936462402, 
42.8951564788818, 43.2648825645447, 43.9498802185059, 44.9562057495117, 
45.7544178432888, 45.4925720214844, 44.8370536804199, 44.1818626403809, 
45.5411309136285, 45.9677772521973, 47.623962020874), PerC.y = c(0.89834948737305, 
1.15055827353951, 1.07068229762836, 1.55762241960512, 0.660824558535851, 
1.49350374350045, 1.30504666418671, 1.22412934748672, 0.899609693326907, 
0.905523389594839, 1.15335080765026, 1.33751682282367, 1.32474641315414, 
1.60474345054736, 0.972492166400552, 0.919991354863598, 1.24135231820059, 
1.57815525005946, 0.997063472019759, 1.15715655760894), n = c(10L, 
10L, 10L, 9L, 10L, 8L, 10L, 10L, 10L, 10L, 8L, 10L, 10L, 9L, 
10L, 10L, 10L, 9L, 10L, 10L)), row.names = c(NA, -20L), class = "data.frame")

And my code:

my.lm.n.s <- lm(d15N.x ~ Year, data = spring.plot)
summary(my.lm.n.s)

my.seg.2.n.s <- selgmented(my.lm.n.s, seg.Z = ~ Year , alpha = 0.05, type = c("bic"),
                         return.fit = TRUE, bonferroni = TRUE, Kmax = 2, msg=TRUE)
my.seg.2.n.s

seg.n.spring <- segmented(my.lm.n.s, seg.Z = ~ Year, npsi = 2)
seg.n.spring

new <- data.frame(c(min(spring.plot$Year), seg.n.spring$psi[, "Est."], max(spring.plot$Year)))
lines <- predict(seg.n.spring, new)

The new object is a data frame so I do not know why it is not working.

My question is very similar to this one.

CodePudding user response：

You have to update one line of code to make it working.

new <- data.frame(Year = c(min(spring.plot$Year), seg.n.spring$psi[, "Est."], max(spring.plot$Year)))

so the data.frame need a proper column name.