This package offers various functions for managing and cleaning data before the application of different approaches. This includes identifying and erasing sudden jumps in dendrometer data not related to environmental change, identifying the time gaps of recordings, and changing the temporal resolution of data to different frequencies. It offers an opportunity for users to use all major approaches in the current state of the art of dendrometer data analysis. In addition, it identifies periods of consecutive days with user-defined climatic conditions in daily meteorological data, then checks how trees responded during that period.
The package requires a well formatted data set as a input. The first column must consist of time in extended date-time format (e.g. yyyy-mm-dd HH:MM:SS) without daylight savings. From the version 0.1.5, the package can handle the data set with extended time format in first column regardless of their format. However, the first column must have datetime from year to second. The dendrometer data for each sensors has to be sorted from the second column onward. The package is flexible to column names but is strict with their order.
Various dendrometers store data in different formats. Some of them generates time in extended date-time format in one column whereas others generate time with different units in different columns. We recommend converting the data to comma separated value (.csv) or plain text (.txt) format before importing to R. However, package can also read excel file from version 0.1.5.
If the dataset looks like in the table 1, one can simply read it using “read.dendrometer” functions. The function is able to read CSV, TXT and EXCEL files.
Table 1: Dendrometer data with extended date-time in one column
#> Registered S3 method overwritten by 'quantmod':
#> method from
#> as.zoo.data.frame zoo| Time | T2 | T3 |
|---|---|---|
| 2016-05-04 18:00:00 | 47.61461 | 35.63046 |
| 2016-05-04 19:00:00 | 47.67029 | 35.67258 |
| 2016-05-04 20:00:00 | 47.76567 | 35.69873 |
| 2016-05-04 21:00:00 | 47.83297 | 35.76167 |
| 2016-05-04 22:00:00 | 47.86202 | 35.77087 |
For plain text (.txt), comma separated (.csv) and excel (.xlsx) format:
df <- read.dendrometer("nepa.txt")If a dataset does not include time in extended date-time format, but separate columns for year, month, day, hour, minute and second (Table 2), it needs additional format after importing to R.
Table 2: Dendrometer data with extended date-time in separate columns
| year | month | day | hours | minutes | seconds | T2 | T3 |
|---|---|---|---|---|---|---|---|
| 2016 | 5 | 4 | 18 | 0 | 0 | 47.61461 | 35.63046 |
| 2016 | 5 | 4 | 19 | 0 | 0 | 47.67029 | 35.67258 |
| 2016 | 5 | 4 | 20 | 0 | 0 | 47.76567 | 35.69873 |
| 2016 | 5 | 4 | 21 | 0 | 0 | 47.83297 | 35.76167 |
| 2016 | 5 | 4 | 22 | 0 | 0 | 47.86202 | 35.77087 |
For comma separated value (.csv) format:
df <- read.csv("nepa2.csv", header = TRUE)
date <- paste(df$year,sprintf('%02d',df$month),sprintf('%02d',df$day), sep="-")
time <- paste(sprintf('%02d',df$hours), sprintf('%02d',df$minutes), sprintf('%02d',df$seconds), sep=":")
datetime <- paste(date, time, sep = " ")
df2 <- data.frame("Time" = datetime)
df2$T2 <- df$T2
df2$T3 <- df$T3For plain text (.txt) format:
df <- read.table("nepa2.txt", header = TRUE)
date <- paste(df$year,sprintf('%02d',df$month),sprintf('%02d',df$day), sep="-")
time <- paste(sprintf('%02d',df$hours), sprintf('%02d',df$minutes), sprintf('%02d',df$seconds), sep=":")
datetime <- paste(date, time, sep = " ")
df2 <- data.frame("Time" = datetime)
df2$T2 <- df$T2
df2$T3 <- df$T3After the formatting, the final dataset must appear similar like in table 3.
Table 3: Dendrometer data after formatting extended date-time in one column
| Time | T2 | T3 |
|---|---|---|
| 2016-05-04 18:00:00 | 47.61461 | 35.63046 |
| 2016-05-04 19:00:00 | 47.67029 | 35.67258 |
| 2016-05-04 20:00:00 | 47.76567 | 35.69873 |
| 2016-05-04 21:00:00 | 47.83297 | 35.76167 |
| 2016-05-04 22:00:00 | 47.86202 | 35.77087 |
Sugam Aryal
Email: sugam.aryal@fau.de