Raised by @sckott,

with enddate = Sys.Date() is there a possibility this will fail if it's a date that the data isn't ready yet? is there at least a good failure mode when that happens?

Trying to run the below code gives an error when the parameter for solar irradiance is defined as "ALL SKY TOA SW DWN"; I get a similar issue with "SI_EF_MIN_OPTIMAL".

In the first case, I get an error message that the "all sky" string is not a valid parameter. In the second, I get an error message that the SI EF min Optimal returns not an energy but an orientation for the panels. In both cases, this is at odds with the documentation on the NASA Power API website. It is possible that these values do not work with the "climatology" temporal average but if that is the case (I have yet to try out others) this is not spelled out clearly in anything I can find.

Thanks!

get_power(community = "SSE", pars = c("TS", "ALLSKY_TOA_SW_DWN"), temporal_average = "CLIMATOLOGY", lonlat = c(averagelon, averagelat))

Hi! Thanks for the package! When creating the files for DSSAT (ICASA format) headers and columns are not in the alignment that the model requires. In the attached file, the first sequence depicts the working structure, and below the structure generated by create_icasa.
How can I adjust these alignments?
Thanks in advance!
new 1.txt

duplicated "data" occurs in REAMDE surface solar energy climatology data data

Repeatedly getting the following error. Nothing seems to be wrong on https://power.larc.nasa.gov/

Error:
A CSV file was not created, this is a server error.
The server may not be responding or is currently responding improperly.
Please check https://power.larc.nasa.gov/ for notifications if
you repeatedly get this error.

Testing with:
power <- get_power(community = "AG",
lonlat = c(-179.5, -89.5),
pars = c("RH2M", "T2M"),
dates = "1985-01-01",
temporal_average = "DAILY")

The arguments for querying Climatology data from POWER can be confusing.

Simplify to use only temporal_average = "CLIMATOLOGY" by setting lonlat = NULL by default and setting lonlat = "GLOBAL" in the get_POWER() function itself when CLIMATOLOGY is detected.

The code below works, but if I change WS10M for WS2M, it returns an error:

""Error:
You have entered an invalid value for community for the supplied pars. One or more pars are not, available for AG, please check.""

get_power(community = "AG",
                         lonlat = c(-49.554,-20.862, -48.562, -19.851),
                         pars = c("PRECTOT", "T2M_MAX", "T2M_MIN", "T2M", "ALLSKY_SFC_SW_DWN", "RH2M", "T2MDEW", "WS10M"), 
                         dates = c("2018-01-01", paste0(Sys.Date())),
                         temporal_average = "DAILY")

Error: 18390 is not a valid entry for date. Enter as YYYY-MM-DD.

In response to Error 18390, could you please add functionality for calling a function to define the "dates =" argument in get_power()?

I would like to be able to use lubridate::today() when I call the dates for get_power() so that I can automatically update my get_power() data on a daily basis.

Example when retrieving data spanning two years.

> cell <- get_cell(lonlat = c(-179.5, -89.5), stdate = "2012-12-30", endate = "2013-01-02")
Downloading: 1.8 kB     

> cell
  YEAR MONTH DAY   YYYYMMDD DOY    LON   LAT    T2M   T2MN   T2MX  RH2M toa_dwn swv_dwn lwv_dwn  DFP2M RAIN
1 2012    12  29 2013-12-29 365 -179.5 -89.5 -28.63 -29.50 -27.74 66.77   47.84   32.40   12.53 -32.83 0.00
2 2012    12  30 2013-12-30 366 -179.5 -89.5 -28.57 -29.62 -27.56 60.89   47.63   39.31   11.02 -33.71 0.00
3 2013    12  30 2012-12-30   1 -179.5 -89.5 -28.10 -29.15 -27.02 64.53   47.81   38.56   10.91 -32.67 0.00
4 2013    12  31 2012-12-31   2 -179.5 -89.5 -27.05 -27.60 -26.29 64.59   46.98   31.54   15.88 -31.66 0.06
  WS10M
1  2.18
2  2.50
3  1.95
4  2.54

Using get_cell() with multiple dates will return only the first date, e.g.,

x <- get_cell(
     lonlat = c(-179.5, 89.5),
     stdate = "2017-1-1",
     endate = "2017-1-3")

will only return data for Jan. 1

Validate user input dates for validity and handle formats other than "YYYY-MM-DD" properly

These are not optional. Leave blank so they fail in a canonical R fashion if not specified

Hi @adamhsparks , when I try to install nasapower, i get the following message and fail to install the package:

library(devtools)
devtools::install_github("adamhsparks/nasapower", build_vignettes = TRUE)

Downloading GitHub repo adamhsparks/nasapower@master
from URL https://api.github.com/repos/adamhsparks/nasapower/zipball/master
Installing nasapower
'/Library/Frameworks/R.framework/Resources/bin/R' --no-site-file --no-environ --no-save
--no-restore --quiet CMD build
'/private/var/folders/f_/hrv_rbld7w74b42yz8psmmv40000gn/T/RtmpFFlUrZ/devtools808a2043affd/adamhsparks-nasapower-63edbdc'
--no-resave-data --no-manual

• checking for file ‘/private/var/folders/f_/hrv_rbld7w74b42yz8psmmv40000gn/T/RtmpFFlUrZ/devtools808a2043affd/adamhsparks-nasapower-63edbdc/DESCRIPTION’ ... OK
• preparing ‘nasapower’:
• checking DESCRIPTION meta-information ... OK
• installing the package to build vignettes
• creating vignettes ... ERROR
  Quitting from lines 66-74 (use-case.Rmd)
  Error: processing vignette 'use-case.Rmd' failed with diagnostics:
  Not Found (HTTP 404).
  Execution halted
  Installation failed: Command failed (1)

Do you have any idea what is going wrong? I've installed devtools already.

The "nasapower" vignette is out of date and should be updated to reflect new functionality and how to create spatial objects using the currently provided functionality.

@adamhsparks I was trying to run the second example but got this error:

get_cell(lonlat = c(-179.5, -89.5), endate = "1983-01-02", vars = "WS10M")

Error in curl::curl_fetch_memory(url, handle = handle) :
The POWER website does not appear to be responding.
Please try again later.

Can't download data files from POWER Data Acces Viewer, i am for a few days now and getting this error "Unable to access data at this time, please check back soon.". I am using the API in python now but is very slow and need faster access to the data.

Thanks.

Raised by @sckott,

seems smart to break apart the GET and content calls on separate lines, w/ a stop_for_status() or similar in between them so users get nice error messages when a server or client error happens

Most software uses lon lat not lat lon.

This will simplify using other R packages like raster that return xy values not yx, etc.

But maybe -999 as well?

Just account for possibility of both being in the data as currently it's set to -999.

As pointed out by @sckott, the examples in function documentation don't run.

Typo locally caused v1.0.5 to be exactly the same as v1.0.4

Also missed out on support for WS2M_MIN, WS2M_MAX and WS2M_RANGE in the release.

https://ropensci.github.io/nasapower/articles/nasapower.html

Hello,
I cannot make 'get_power' working for some parameters like "SI_EF_OPTIMAL". They can be accessed using NASA's web interface:
https://power.larc.nasa.gov/downloads/POWER_SinglePoint_Climatology_033d04N_093d00E_8ddb2015.txt
which creates this order:
https://power.larc.nasa.gov/cgi-bin/v1/DataAccess.py?&request=execute&identifier=SinglePoint&parameters=SI_EF_TILTED_SURFACE&userCommunity=SSE&tempAverage=CLIMATOLOGY&outputList=ASCII&lat=33.4000&lon=93.0000
Here is the code I use:

sol_daily_BJ <- get_power(community = "SSE",
                          lonlat = c(93, 33.4),
                          pars = c("SI_EF_OPTIMAL"),
                          temporal_average = "CLIMATOLOGY"
)

Which returns:
"Error:
Something went wrong with the query, no data were returned.
Please see https://power.larc.nasa.gov for potential server issues."
I'm not sure if this is the package or the POWER issue, or I'm just writing it wrong way.
Here is my system info:

devtools::session_info()

• Session info --------------------------------------------------------------------
  setting value
  version R version 3.5.3 (2019-03-11)
  os Windows >= 8 x64
  system x86_64, mingw32
  ui RStudio
  language (EN)
  collate English_United States.1252
  ctype English_United States.1252
  tz America/New_York
  date 2019-04-05

Using rvest? import parameters from https://power.larc.nasa.gov/docs/v1/ and document using data-raw for reproducibility and easy updates if there are changes to data set.

rOpenSci guidelines suggest an example in README and that if there is code it be generated from a .Rmd file.

@sckott wrote,

Also about get_power. it's not a good idea to have a required parameter (w/o a default) after a non-required parameter. that is, you have function(community, lonlat, pars, dates = NULL, temporal_average), whereas function(community, lonlat, pars, temporal_average, dates = NULL) would make more sense b/c all required paramters are first, then the last is optional

If the server/web-page is not responding the package will continuously try to retrieve the data for several minutes until failing.

Add function to check server response before requesting data.

Using get_region() with multiple dates will return only the first date, e.g.,

oz <- get_region(lonlat = c(112.91972, 159.256088, -55.11694, -9.221099),
                    stdate = "2017-1-1", endate = "2017-1-2")

will only return data for Jan. 1 for Australia.

I do this post for ask you about the main site https://power.larc.nasa.gov/ because, it doesn't work and i want know if you kwnow about this... thanks

If meta is not logical, stop execution and alert user.

The POWER team are migrating to a new API by 2021-07-01, https://power.larc.nasa.gov/beta/docs/services/api/

v4.0 will take advantage of this new functionality.

• Implement new API V2 calls
• Implement new functions to take advantage of the new manager API providing information on parameters
• Consolidate, standardise and optimise functions that may be shared between get_power() and the new query_parameters()
• Write tests for new and revised functions
• Update vignette to reflect new functionality of nasapower and the POWER API V2
• Check functionality against the new API, e.g. rate limits as reported in #52 (comment)

The vignette shows this map for get_region() example.

However, this is the current result.

Add a summary table of missing data for ICASA and MET files to alert users when saving files.

I just followed the example provided in the package to access daily data, see below.

library(nasapower)
daily_single_ag <- get_power(community = "AG",
                             lonlat = c(151.81, -27.48),
                             pars = c("RH2M", "T2M", "PRECTOT"),
                             dates = c("1985-01-01", "1985-01-31"),
                             temporal_average = "DAILY"
)

daily_single_ag

It through the error below

Error: 
Something went wrong with the query, no data were returned. Please see <https://power.larc.nasa.gov> for potential server issues.

From @sckott,

instead of this pattern https://github.com/adamhsparks/nasapower/blob/master/R/get_cell.R#L111-L135 can do instead list(lon = longlat[1], lat = lonlat[2], ...), etc. and pass to GET as first param just the base URL, and pass that list to query

Hello,

I am a developer with the NASA POWER Project and would like to let you know that there is a new link for the parameters.json file. We had to change hosing environments and that required some changes to the URL's structure.

Please use the following link in your application:

• https://power.larc.nasa.gov/system/parameters.json

Replacing:

• https://power.larc.nasa.gov/RADAPP/GEODATA/powerWeb/POWER_Parameters_v110.json

Other Links:

• https://power.larc.nasa.gov/docs/
• https://power.larc.nasa.gov/docs/methodology/

Calls to the API all day are returning the error below:

Error: Something went wrong with the query, no data were returned. Please see <https://power.larc.nasa.gov> for potential server issues.

From @sckott's review:

You could be a bit more detailed on your parameters man file, e.g., A list with 141 objects. What kind of objects? And it looks like the list is named by something useful?

Use the APSIM package to generate a .met file for use in APSIM using POWER data.

e.g. create_met <- (years = NULL, latlon = NULL) and the function generates a full APSIM .met file object based only on date(s) and location(s) with appropriate data.

I am trying to run these lines of code (provided as example on readme.md).

if (!require(devtools)) {
  install.packages("devtools")
}
devtools::install_github("ropensci/nasapower",
                         build_vignettes = TRUE)

#install.packages("nasapower",dep = T)
library(nasapower)

daily_ag <- get_power(community = "AG",
                      lonlat = c(151.81, -27.48),
                      pars = c("RH2M", "T2M", "PRECTOT"),
                      dates = "1985-01-01",
                      temporal_average = "DAILY")

However, after a minute (or more) of waiting, i click on the stop button and the following returns:

Error in curl::curl_fetch_memory(x$url$url, handle = x$url$handle) :
Operation was aborted by an application callback

I have been running these same lines of code before and it was working flawlessly, but since yesterday i get these error returns. There is any server instability these days?

Thomas Leeper has created a package to read/write CSVY files (CSV with a YAML header).

This could be useful for storing the metadata that the POWER API serves and is currently only displayed in the header information but not exported from the R session unless using a .Rds or other R file to save the data.

https://github.com/leeper/csvy

For the interannual time-series, only year is necessary for start/end date.

For climatology it's not used.

Need to include checks for these conditions and include restrictions and messages.

Create new function to retrieve ICASA formatted data for DSSAT and create an S4 object in R and optionally write it to disk

Raised by @sckott

get_cell(lonlat = c(-179.5, -89.5), vars = 'asdfasdf')
#> Error in `[.data.frame`(NASA, , c(refcols, setdiff(names(NASA), refcols))) :
#>  undefined columns selected