R/shift.R

In benmbutler/powdR: Full Pattern Summation of X-Ray Powder Diffraction Data

#This shifting function has deprecated.

# .shift <- function(smpl, lib, max_shift, x, res, obj) {
#
#   #create a blank list
#   pure_patterns <- list()
#
#   #Create a new sample pattern with res times the resolution of the original
#   sample_pattern <- data.frame(stats::approx(x = smpl[,1], y = smpl[,2],
#                                              method = "linear", n = length(smpl[,1]) * res))
#
#   TTH_long <- sample_pattern[[1]]
#   sample_long <- sample_pattern[[2]]
#
#   #Do the same for all data in the reference library
#   for (i in 1:ncol(lib$xrd)) {
#     pure_patterns[[i]] <- stats::approx(x = lib$tth, y = lib$xrd[, i],
#                                         method = "linear", n = nrow(lib$xrd) * res)[[2]]
#   }
#
#   #convert from list to data frame, to matrix
#   if (length(pure_patterns) == 1) {
#     pure_patterns <- data.frame("phase" = pure_patterns[[1]])
#     names(pure_patterns) <- names(lib$xrd)
#     pure_patterns <- as.matrix(pure_patterns)
#   } else {
#     pure_patterns <- data.frame(pure_patterns)
#     names(pure_patterns) <- names(lib$xrd)
#     pure_patterns <- as.matrix(pure_patterns)
#   }
#
#   #Define a value that will be used to shift the data.
#   TTH_res <- (TTH_long[length(TTH_long)] - TTH_long[1])/(length(TTH_long)-1)
#   #Round up the number of increments the data can shift by (based on max_shift)
#   shift_value <- round(max_shift/TTH_res)
#
#   #Shorten the sample pattern and 2theta to account for
#   #the maximum/minimum shifts that might be applied
#   sample_long <- sample_long[((shift_value + 1):(length(sample_long)-shift_value))]
#   TTH_long <- TTH_long[((shift_value + 1):(length(TTH_long) - shift_value))]
#
#   #define an integer sequence of positive and negative shifts
#   initial_shift <- c((0 - shift_value):shift_value)
#
#   shifting_length <- nrow(pure_patterns)-shift_value
#
#   #Create a matrix of the shortened length that will be used during alignment
#   shift_mat <- pure_patterns[((shift_value + 1):(shifting_length)), ]
#
#   if (ncol(pure_patterns) == 1) {
#     shift_mat <- data.frame("phase" = shift_mat)
#     names(shift_mat) <- names(lib$xrd)
#     shift_mat <- as.matrix(shift_mat)
#   }
#
#   #define blank lists to be populated during alignment
#   v <- list()
#   vm <- list()
#   vf <- list()
#   d <- list()
#
#   #This vector will be used to identify the optimum shift
#   #dmin <- c()
#
#   #This vs matrix will be populated with the aligned patterns
#   vs <- shift_mat
#
#   cat("\n-Shifting patterns")
#
#   for (i in 1:ncol(shift_mat)) {
#     for (j in 1:length(initial_shift)) {
#
#       v[[j]] <- pure_patterns[c(((shift_value + 1) + (initial_shift[j])):(shifting_length + (initial_shift[j]))), i]
#
#       #adjusted matrix for each shift
#
#       vm[[j]] <- shift_mat
#       # #add the shifted data
#       vm[[j]][,i] <- v[[j]]
#
#       # compute the fitted pattern
#       vf[[j]] <- apply(sweep(vm[[j]], 2, x, "*"), 1, sum)
#
#       #Compute Rwp if defined in obj
#       if (obj == "Rwp") {
#
#       d[[j]] <- sqrt(sum((1/sample_long) * ((sample_long - vf[[j]])^2)) /
#                        sum((1/sample_long) * (sample_long^2)))
#
#       }
#
#       #Compute the Delta is defined in obj
#       if (obj == "Delta") {
#
#       d[[j]] <- sum(abs(sample_long - vf[[j]]))
#
#       }
#
#       #Compute R if defined in obj
#       if (obj == "R") {
#
#         d[[j]] <- sqrt(sum((sample_long - vf[[j]])^2)/sum(sample_long^2))
#
#       }
#
#       #identify which shifted pattern results in minimum Delta
#       #dmin[[i]] <- which.min(d)
#
#       #Populate a library with the optimumly shifted references
#       vs[, i] <- vm[[which.min(d)]][, i]
#     }
#   }
#
#   #re-approximate the data to the old TTH resolution (i.e. reduce by res times)
#   vs_short <- list()
#
#   #re-approximate the reference library
#   for (i in 1:ncol(vs)) {
#     vs_short[[i]] <- stats::approx(x = 1:nrow(vs), y = vs[ , i], method = "linear", n = (nrow(vs) / res))[[2]]
#   }
#   #Convert from list to data frame to matrix
#
#   if (length(vs_short) == 1) {
#     vs_short <- data.frame("phase" = vs_short[[1]])
#     names(vs_short) <- names(data.frame(vs))
#     vs_short <- as.matrix(vs_short)
#   } else {
#     vs_short <- data.frame(vs_short)
#     names(vs_short) <- names(data.frame(vs))
#     vs_short <- as.matrix(vs_short)
#   }
#
#   #reapproximate the sample
#   sample_pattern <- stats::approx(x = TTH_long, y = sample_long, method = "linear", n = (nrow(vs) / res))[[2]]
#
#   #reapproximate the 2theta
#   TTH_short <- stats::approx(x = TTH_long, y = vs[, 1], method = "linear", n = (nrow(vs) / res))[[1]]
#
#   vs <- vs_short
#   TTH <- TTH_short
#
#   out <- list("smpl" = data.frame("tth" = TTH, "counts" = sample_pattern),
#               "lib" = vs)
#
# }