A loess regression is fit to the surface measurements and residuals are
calculated.
The most extreme 0.25% of residuals are filtered from further consideration.
The result is called the raw signature of the bullet land.
Adapted from fit_loess in bulletr
Arguments
- ccdata
The crosscut as returned from x3p_to_df, grooves need to be removed ahead of time
- span
The span to use for the loess regression
Value
a list of a data frame of the original bullet measurements extended by loess fit, residuals, and standard errors and two plots: a plot of the fit, and a plot of the bullet's land signature.
Examples
library(dplyr)
ccdata <- data_frame(
x = seq(0, 6, .001),
value = 10 - (3 - x)^2 + rnorm(length(x), sd = .25)
)
cc_fit_loess(ccdata = ccdata)
#> # A tibble: 6,001 × 7
#> x value fitted raw_sig se abs_resid chop
#> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <lgl>
#> 1 0 0.462 1.01 -0.549 0.0123 0.549 FALSE
#> 2 0.001 0.788 1.02 -0.228 0.0123 0.228 FALSE
#> 3 0.002 0.997 1.02 -0.0259 0.0123 0.0259 FALSE
#> 4 0.003 0.567 1.03 -0.462 0.0123 0.462 FALSE
#> 5 0.004 1.30 1.03 0.270 0.0122 0.270 FALSE
#> 6 0.005 0.764 1.04 -0.277 0.0122 0.277 FALSE
#> 7 0.006 0.824 1.05 -0.222 0.0122 0.222 FALSE
#> 8 0.007 0.842 1.05 -0.211 0.0122 0.211 FALSE
#> 9 0.008 1.06 1.06 -0.00254 0.0122 0.00254 FALSE
#> 10 0.009 0.733 1.06 -0.331 0.0122 0.331 FALSE
#> # ℹ 5,991 more rows