In m-saenger/metar: Parse METAR reports

knitr::opts_hooks$set(eval = function(options) {
  if (options$engine %in% c("bash", "python")) {
    options$eval <- FALSE
  }
  options
})

Set up R environment

Latest METAR R libraries are installed on /prod enviroment. Note that libraries built under R 3.X.X under possibly not compatible with those built under R 4.X.X. Make sure server runs R > 4.0.0.

Test R version ``` {bash eval=FALSE}

R --version

### Modify local R profile 

Makes sure R loads libraries from a specific path. R reads user `.Renviron` file upon start, thus the location of the libraries can be specified here. Caution: if Python script is run by some `sysuser` the according `.Renviron` file in the home dir needs be modifed.

``` {bash eval=FALSE}
nano ~/.Renviron

R_LIBS=prod/...

Check whether R profile works

Open R in shell ``` {bash eval=FALSE}

R

``` {R eval=FALSE}
> .libPaths() # List paths where R looks for libraries
[1] "C:/R/lib"                           "C:/Program Files/R/R-4.1.1/library"
> library(metar.parsing) # Check whether library metar.parsing can be loaded
> parse_metar_cld("FEW 005") # Test library
# Should return table of parsed cloud group
> quit("no") # Close R

Now R (and rpy2) should load libraries from the correct path.

Set up Python

rpy2

R code can be run with rpy2 framework: https://rpy2.github.io/doc/latest/html/index.html

sudo pip3 install rpy2

Demo script

``` {Python eval=FALSE}

import rpy2.robjects as robjects import pandas as pd

from rpy2.robjects.packages import importr from rpy2.robjects import pandas2ri from rpy2.robjects.conversion import localconverter

Import METAR libraries

metar_parser = importr("metar.parsing") metar_analytics = importr("metar.analytics")

Import cloud parser from R

parse_metar_cld = robjects.r('parse_metar_cld')

sample data

clouds_autometar = ["FEW030", "BKN070", "SCT050", "FEW080", "SCT299CB BKN300"]

Convert to R character vector

r_clouds_autometar = robjects.StrVector(clouds_autometar)

Call R cloud parser

r_result = parse_metar_cld(r_clouds_autometar) print(r_result)

Convert to pandas

with localconverter(robjects.default_converter + pandas2ri.converter): py_result = robjects.conversion.rpy2py(r_result)

print(py_result)

```

m-saenger/metar documentation built on Jan. 29, 2025, 12:06 a.m.