CHCN (version 1.5)

CHCN-package: A package for Building Historical Climate data for Canada

Description

The package provides tools for scraping climate data from the Environment Canada website and turning it into R objects that can easily be used by analysts. The functions provided allow the user to download a master list of all the data at Environment Canada and to build a collection of monthly climate data from those resources. The master list is downloaded and then the stations which provide monthly data is extracted. Those station Ids are used to make http requests and the data is downloaded as csv files. There are function provided to parse the local csv files and create station "Inventories" with limited metadata ( station name, latitude, longitude, etc). In addition, functions for selecting and compiling temperature data are provided. Other data, such as snowfall, rainfall, can also be easily extracted with simple R commands. Data can be reformated into a format similar to that provided by NOAA's GHCN ( Global Historical Climate Network) and integration with the package RghcnV3 is trivial.

key functions and building the database

The process of building the database is shown in the demo files. The process starts by downloading the master station list. The format of this list has changed somewhat since the first publication of this code, so version prior to 1.3 will be broken by that change. The first call to make is downloadMaster() That call creates a master list of all stations and their web ids. If that download fails there is a backup version of the file shipped with the package. It can be read using the function readLocalMaster. See that documentation for directions. Next, we create a list of monthly stations from that master list using Stations <- writeMonthlyStations() This writes a file of the stations that report monthly. Then we scrape the website: scrapeToCsv(Stations), passing that function our list of monthly stations. When this process completes we check for missing or empty csv files: EMPTY <- getEmptyCsv(). If the list is null then we have no empty downloads. Checking for missing downloads is accomplished by getMissingScrape. Assuming that all files are downloaded, then we can proceed to create datasets and inventories. data <- createDataset() will create a dataset by reading all the csv files. As there is more than temperature data here, we want to save it all: writeData(data). Next, we create an inventory of all the stations: inv <- createInventory(). This inventory can be saved with a simple write.table. When we save this we preserve all the orginal metadata in the format we downloaded it in: write.table(inv,"masterInventory.inv"). Next for working with other packages ( like RghcnV3) we want to save data in a friendly format: writeInventory(inv) will write a GHCN style inventory with variables named appropriately and put into the correct columns for RghcnV3. Lastly, we want to select certain data from the master datafile and write it out. To select Tmean we do the following Mean <- formatGhcn(data, dataColumn = 7). Data column 7 ( use colnames(data) to see the entire list of options ). lastly, we write out the data in ghcn format writeGhcn(data, directory = DATA.DIRECTORY, filename = "TaveCHCN.dat")

Details

Package:

CHCN

Type:

Package

Version:

1.5

Date:

2012-06-07

License:

GPL (>= 2)

LazyLoad:

yes

LazyData:

FALSE

References

http://climate.weatheroffice.gc.ca

Examples

## Not run: 
#      downloadMaster()
#      Stations <- writeMonthlyStations()
#      scrapeToCsv(Stations)
#      EMPTY <- getEmptyCsv()
#      if (is.null(EMPTY)){
#          data <- createDataset()
#          ###  save all the data
#          writeData(data)
#          inv  <- createInventory()
#          write.table(inv,"masterInventory.inv")
#          # write a ghcn style inventory
#          writeInventory(inv)
#          # select Tave data
#          Mean <- formatGhcn(data, dataColumn = 7)
#          writeGhcn(Mean)
#  
# } else{
#   scrapeToCsv(EMPTY)
# }
# 
# ## End(Not run)