returns density of convolution of each component of a normal mixture with N(0,s^2) at x. Note that convolution of two normals is normal, so it works that way
# S3 method for normalmix
comp_dens_conv(m, data, ...)
a k by n matrix
mixture distribution with k components
a list with components x and s to be interpreted as a normally-distributed observation and its standard error
other arguments (unused)