Tools for Hystra files. See https://tonyladson.wordpress.com/2015/05/06/reading-in-data-from-supplied-from-hydstra/

HydstraHelper.R

require(ggplot2)
require(zoo)
require(dplyr)


# MakeNames_Hydstra
# Hydstra files have names spread across 4 rows. This gist, turns the data into a single string

MakeNames_Hydstra <- function(Hydstra.csv){
  
  r1 <- as.vector(unlist(Hydstra.csv[1, ]))
  r1 <- zoo::na.locf(r1)
  r1 <- paste0('X', r1)
  
  r3 <- as.vector(unlist(Hydstra.csv[3, ]))
  r3 <- na.locf(r3)
  r3 <- tolower(substring(r3,1,1))
  
  r4 <- as.vector(unlist(Hydstra.csv[4, ]))
  r4 <- tolower(substring(r4,1,1))
  
  my.names <- paste(r1, r3, r4, sep=".")
  my.names[1] <- "date.time"
  my.names
}

# HydstraSummary

# produce a data frame that lists column name, start of non-missing data, end of non-missing data
# number of data points, number of missing and percentage missing

# date.col = name of the column that contains dates
# df = data frame

HydstraSummary <- function(date.col, df) {
  
  
  .HydraColumnSummary <- function(my.col, date.col, df){
    start.row <- min(which(!is.na(df[[my.col]])))
    end.row <- max(which(!is.na(df[[my.col]])))
    if(is.finite(start.row) & is.finite(end.row)) {
      n.missing <- sum(is.na(df[[my.col]][start.row:end.row]))
      n.missing.pc <- round(100* mean(is.na(df[[my.col]][start.row:end.row])),2)
      n = end.row - start.row + 1
      start <- df[[date.col]][min(which(!is.na(df[[my.col]])))]
      end   <- df[[date.col]][max(which(!is.na(df[[my.col]])))]
    } else {
      n.missing = NA
      n.missing.pc = 100
      n = 0
      start = NA
     end = NA
    
    }
    data.frame(my.col = my.col, start, end, n, n.missing, n.missing.pc)
  
  }

  x <-  lapply(names(df), .HydraColumnSummary, date.col = date.col, df=df)
  out <- do.call(rbind,x)
  out
}


#  Given the file name of a Hydstra file, returns a data frame

Make_df_Hydstra <- function(fname) {
  
  # Read header
  hydstra.head <- read.csv(file = fname, header = FALSE, nrows = 4, na.strings=c(NA,""), stringsAsFactors = FALSE)
  
  # Read data
  hydstra.data <- tbl_df(read.csv(file = fname, header = FALSE, nrows = -1, skip = 4, na.strings=c(NA,""), stringsAsFactors = FALSE))
  
  # Assign names
  my.names <- MakeNames_Hydstra(hydstra.head)
  my.names[1] <- 'date.time'
  names(hydstra.data) <- my.names
  
  # return data frame
  
  hydstra.data
  
}