There are a number of errors that you might encounter using thisrATTAINS. Here is a list of potential errors and fixes. Feel free toraise an issue if I missed something.
Network Connectivity
The following error message likely indicates an issue connecting tothe EPA server:
state_summary(organization_id="TCEQMAIN", reporting_cycle="2022")Potential issues/fixes:
- Check your network connection.
- Checkattains.epa.gov. If you are ableto connect, a warning notice about accessing U.S. Government informationsystems should show in your web browser.
- Occasionally proxy systems used in corporate IT systems cause issueswith connections (see:https://stackoverflow.com/questions/59796178/r-curlhas-internet-false-even-though-there-are-internet-connection).I’ve tried to account for this in the package, but you might run intooccasional issues.
Server Response
The server might also return http code messages. The most common willbe 404 or 429. rATTAINS will generally provide a simple message anderror when this is encountered:
actions(action_id="R8-ND-2018-03")#>#> Error: parse error: premature EOF#>#> (right here) ------^Potential issues/fixes:
- Wait until the server is responsive.
- Make less frequent requests.
Parsing Errors
The default behavior in rATTAINS is to parse JSON data downloadedfrom the API to one or more dataframes. These are returned as a singledataframe or list of dataframes depending on the function. rATTAINS alsotries to flatten the data as much as possible. This design choicemight have been a mistake because it can become asource of errors if the data returned by the API changes or isinconsistent. As of version 1.0.0 of the package the.unnest argument was added to most functions. By setting.unnest=FALSE many of these problems should be avoided.
Default behavior:
state_summary(organization_id="TDECWR", reporting_cycle="2022")#> Unable to further unnest data, check for nested dataframes.#> $items#># A tibble: 20 × 18#> organizationIdentifier organizationName organizationTypeText reportingCycle#><chr><chr><chr><chr>#> 1 TDECWR Tennessee State 2022#> 2 TDECWR Tennessee State 2022#> 3 TDECWR Tennessee State 2022#> 4 TDECWR Tennessee State 2022#> 5 TDECWR Tennessee State 2022#> 6 TDECWR Tennessee State 2022#> 7 TDECWR Tennessee State 2022#> 8 TDECWR Tennessee State 2022#> 9 TDECWR Tennessee State 2022#>10 TDECWR Tennessee State 2022#>11 TDECWR Tennessee State 2022#>12 TDECWR Tennessee State 2022#>13 TDECWR Tennessee State 2022#>14 TDECWR Tennessee State 2022#>15 TDECWR Tennessee State 2022#>16 TDECWR Tennessee State 2022#>17 TDECWR Tennessee State 2022#>18 TDECWR Tennessee State 2022#>19 TDECWR Tennessee State 2022#>20 TDECWR Tennessee State 2022#># ℹ 14 more variables: cycleStatus <chr>, combinedCycles <list>,#># waterTypeCode <chr>, unitsCode <chr>, useName <chr>,#># `Fully Supporting` <dbl>, `Fully Supporting-count` <int>,#># `Not Assessed` <dbl>, `Not Assessed-count` <int>, parameters <list>,#># `Not Supporting` <dbl>, `Not Supporting-count` <int>,#># `Insufficient Information` <dbl>, `Insufficient Information-count` <int>Using.unnest=FALSE returns nested columns. The tidyrfamily ofunnest() functions is an easy way to flatten thisdata:
df<-state_summary( organization_id="TDECWR", reporting_cycle="2022", .unnest=FALSE)df$items|>dplyr::select(parameters)|>tidyr::unnest_wider(parameters)|>tidyr::unnest(c(parameterGroup,Cause,"Cause-count","Meeting Criteria","Meeting Criteria-count","Insufficient Information","Insufficient Information-count"))#># A tibble: 67 × 7#> parameterGroup Cause `Cause-count` `Meeting Criteria`#><chr><dbl><int><dbl>#> 1 NUTRIENTS29134. 3NA#> 2 SALINITY/TOTAL DISSOLVED SOLIDS/CHL… 56.1 1NA#> 3 PH/ACIDITY/CAUSTIC CONDITIONS23051 1NA#> 4 SALINITY/TOTAL DISSOLVED SOLIDS/CHL… 56.1 1NA#> 5 PH/ACIDITY/CAUSTIC CONDITIONS23107. 2NA#> 6 ORGANIC ENRICHMENT/OXYGEN DEPLETION5269. 5NA#> 7 SEDIMENT3772. 7NA#> 8 SALINITY/TOTAL DISSOLVED SOLIDS/CHL… 56.1 1NA#> 9 AMMONIA 56.1 1NA#>10 TEMPERATURENANA20459#># ℹ 57 more rows#># ℹ 3 more variables: `Meeting Criteria-count` <int>,#># `Insufficient Information` <dbl>, `Insufficient Information-count` <int>If the above option doesn’t work, rATTAINS can also provide the rawJSON data from the API. Thejsonlite 📦provides tools to convert JSON to nested lists and tibbles prettyeasily. First, use thetidy=FALSE argument to return theunparsed JSON string, then uses jsonlite to convert that data to anested list, then use tidyr to access the nested dataframes!
raw_data<-state_summary( organization_id="TDECWR", reporting_cycle="2022", tidy=FALSE)list_data<-jsonlite::fromJSON(raw_data, simplifyVector=TRUE, simplifyDataFrame=TRUE, flatten=FALSE)df<-tibble::as_tibble(list_data$data)df|>tidyr::unnest(reportingCycles)|>tidyr::unnest(waterTypes)|>tidyr::unnest(useAttainments)#># A tibble: 20 × 18#> organizationIdentifier organizationName organizationTypeText reportingCycle#><chr><chr><chr><chr>#> 1 TDECWR Tennessee State 2022#> 2 TDECWR Tennessee State 2022#> 3 TDECWR Tennessee State 2022#> 4 TDECWR Tennessee State 2022#> 5 TDECWR Tennessee State 2022#> 6 TDECWR Tennessee State 2022#> 7 TDECWR Tennessee State 2022#> 8 TDECWR Tennessee State 2022#> 9 TDECWR Tennessee State 2022#>10 TDECWR Tennessee State 2022#>11 TDECWR Tennessee State 2022#>12 TDECWR Tennessee State 2022#>13 TDECWR Tennessee State 2022#>14 TDECWR Tennessee State 2022#>15 TDECWR Tennessee State 2022#>16 TDECWR Tennessee State 2022#>17 TDECWR Tennessee State 2022#>18 TDECWR Tennessee State 2022#>19 TDECWR Tennessee State 2022#>20 TDECWR Tennessee State 2022#># ℹ 14 more variables: cycleStatus <chr>, combinedCycles <list>,#># waterTypeCode <chr>, unitsCode <chr>, useName <chr>,#># `Fully Supporting` <dbl>, `Fully Supporting-count` <int>,#># `Not Assessed` <dbl>, `Not Assessed-count` <int>, parameters <list>,#># `Not Supporting` <dbl>, `Not Supporting-count` <int>,#># `Insufficient Information` <dbl>, `Insufficient Information-count` <int>