Bug in query.trait.data in relation to FRC_LC trait #2269

serbinsh · 2019-02-01T15:02:59Z

Bug Description

I have hit an error when running query.trait.data() with trait "FRC_LC."

This warning is displayed once per session.
Error in if (nrow(result) == 0) { : argument is of length zero
> traceback()
12: query.trait.data(trait = trait, spstr = PEcAn.utils::vecpaste(ids),
        con = con, update.check.only = update.check.only, ids_are_cultivars = ids_are_cultivars)
11: FUN(X[[i]], ...)
10: lapply(traits$name, function(trait) {
        query.trait.data(trait = trait, spstr = PEcAn.utils::vecpaste(ids),
            con = con, update.check.only = update.check.only, ids_are_cultivars = ids_are_cultivars)
    })
9: PEcAn.DB::query.traits(ids = pft_members$id, priors = traits,
       con = dbcon, update.check.only = TRUE, ids_are_cultivars = (pfttype ==
           "cultivar"))
8: FUN(X[[i]], ...)
7: lapply(pfts, get.trait.data.pft, modeltype = modeltype, dbfiles = dbfiles,
       dbcon = dbcon, forceupdate = forceupdate, trait.names = trait.names)
6: PEcAn.DB::get.trait.data(pfts = pfts, modeltype = modeltype,
       dbfiles = dbfiles, database = database, forceupdate = forceupdate)
5: PEcAn.workflow::runModule.get.trait.data(settings) at workflow.R#76
4: eval(ei, envir)
3: eval(ei, envir)
2: withVisible(eval(ei, envir))
1: source("workflow.R")
>

Specific to this line in the function:

pecan/base/db/R/query.trait.data.R

Line 541 in 7e7b70a

if (nrow(result) == 0) {

> query.trait.data("FRC_LC", "30", con = con)
2019-02-01 09:50:42 INFO   [query.trait.data] :
   ---------------------------------------------------------
2019-02-01 09:50:42 INFO   [query.trait.data] : FRC_LC
Error in postgresqlExecStatement(conn, statement, ...) :
  RS-DBI driver: (could not Retrieve the result : ERROR:  invalid input syntax for integer: ""
LINE 1: ...riables.id = covariates.variable_id where trait_id in ( '' )
                                                                   ^
)
2019-02-01 09:50:42 SEVERE [db.query] :
   Error executing db query 'select covariates.trait_id,
   covariates.level,variables.name from covariates left join variables on
   variables.id = covariates.variable_id where trait_id in ( '' )'
   errorcode=0 message='ERROR: invalid input syntax for integer: "" LINE 1:
   ...riables.id = covariates.variable_id where trait_id in ( '' ) ^ '
Error in PEcAn.logger::logger.severe(paste("Error executing db query '",  :
  Error executing db query 'select covariates.trait_id, covariates.level,variables.name from covariates left join variables on variables.id = covariates.variable_id where trait_id in ( '' )' errorcode=0 message='ERROR:  invalid input syntax for integer: ""
LINE 1: ...riables.id = covariates.variable_id where trait_id in ( '' )
                                                                   ^
'
In addition: Warning message:
In postgresqlQuickSQL(conn, statement, ...) :
  Could not create execute: select covariates.trait_id, covariates.level,variables.name from covariates left join variables on variables.id = covariates.variable_id where trait_id in ( '' )

But for other traits,

> query.trait.data("Vcmax", "938", con = con)
2019-02-01 09:58:00 INFO   [query.trait.data] :
   ---------------------------------------------------------
2019-02-01 09:58:00 INFO   [query.trait.data] : Vcmax
2019-02-01 09:58:00 INFO   [query.trait.data] : Median Vcmax : 21.73
2019-02-01 09:58:00 INFO   [query.trait.data] :
   ---------------------------------------------------------
     id citation_id site_id treatment_id    name       date     time
7  2638          62     368          498 Control 2009-06-16 00:00:00
8  2639          62     368          498 Control 2009-07-24 00:00:00
9  2640          62     368          498 Control 2009-08-28 00:00:00
10 2641          62     368          498 Control 2009-09-18 00:00:00
11 2642          62     368          499   Nfert 2009-06-16 00:00:00
12 2643          62     368          499   Nfert 2009-07-24 00:00:00
13 2644          62     368          499   Nfert 2009-08-28 00:00:00
14 2645          62     368          499   Nfert 2009-09-18 00:00:00
   cultivar_id specie_id  mean statname  stat n vname month      lon     lat
7           19       938 25.62       SE  8.75 4 Vcmax     6 -88.2373 40.0415
8           19       938 23.05       SE  4.17 4 Vcmax     7 -88.2373 40.0415
9           19       938 12.61       SE  6.81 4 Vcmax     8 -88.2373 40.0415
10          19       938 12.48       SE  9.89 4 Vcmax     9 -88.2373 40.0415
11          19       938 29.43       SE 10.60 4 Vcmax     6 -88.2373 40.0415
12          19       938 20.41       SE  8.28 4 Vcmax     7 -88.2373 40.0415
13          19       938 23.91       SE 11.85 4 Vcmax     8 -88.2373 40.0415
14          19       938  8.83       SE  3.02 4 Vcmax     9 -88.2373 40.0415
   control greenhouse mean_unconverted stat_unconverted
7     TRUE      FALSE            25.62             8.75
8     TRUE      FALSE            23.05             4.17
9     TRUE      FALSE            12.61             6.81
10    TRUE      FALSE            12.48             9.89
11   FALSE      FALSE            29.43            10.60
12   FALSE      FALSE            20.41             8.28
13   FALSE      FALSE            23.91            11.85
14   FALSE      FALSE             8.83             3.02

All good.

Looking at "blame" I think it has to do with this change from 11 months ago: #1826

And perhaps others haven't run into this since not many PFTs actually have FRC data available. It just so happens that my new pine barrens PFT does.

To Reproduce

Steps to reproduce the behavior:
query.trait.data("FRC_LC", "30", con = con)
https://modex.bnl.gov/bety/species/30 ; Acer rubrum

seems to only be related to Acer rubrum, as far as I can tell

BUT

others DO work, e.g.
query.trait.data("FRC_LC", "32", con = con)

query.trait.data("FRC_LC", "1130", con = con)

The text was updated successfully, but these errors were encountered:

serbinsh · 2019-02-01T15:05:06Z

Possible that PR is a red herring, though I suspect there is something up with the species record clashing with that cultivar PR. @dlebauer @infotroph any thoughts?

serbinsh · 2019-02-01T15:33:49Z

OK, hmm...

> query.trait.data("FRC_LC", PEcAn.utils::vecpaste(c("30","1014","1017","1130","1142","1170","1178","9303","9315")),
+                  con = con, update.check.only = FALSE, ids_are_cultivars = FALSE)
2019-02-01 10:33:05 INFO   [query.trait.data] :
   ---------------------------------------------------------
2019-02-01 10:33:05 INFO   [query.trait.data] : FRC_LC
2019-02-01 10:33:05 INFO   [query.trait.data] : Median FRC_LC : 1.33
2019-02-01 10:33:05 INFO   [query.trait.data] :
   ---------------------------------------------------------
     id citation_id site_id treatment_id name date     time cultivar_id
1 26702         336      NA            0 <NA> <NA> 00:00:00           0
2 26718         336      NA            0 <NA> <NA> 00:00:00           0
3 26719         336      NA            0 <NA> <NA> 00:00:00           0
4 26726         336      NA            0 <NA> <NA> 00:00:00           0
  specie_id  mean statname stat n  vname month lon lat control greenhouse
1      1017 0.994     <NA>   NA 1 FRC_LC    NA  NA  NA      NA         NA
2      1178 1.590     <NA>   NA 1 FRC_LC    NA  NA  NA      NA         NA
3      1130 1.270     <NA>   NA 1 FRC_LC    NA  NA  NA      NA         NA
4      1170 1.390     <NA>   NA 1 FRC_LC    NA  NA  NA      NA         NA
  mean_unconverted stat_unconverted
1            0.994               NA
2            1.590               NA
3            1.270               NA
4            1.390               NA

now I am not so sure what the issue is and why the call to query traits returns an error?

serbinsh · 2019-02-01T15:37:49Z

Aha!

> data <- query.trait.data("stem_respiration_rate", PEcAn.utils::vecpaste(c("30","1014","1017","1130","1142","1170","1178","9303","9315")),
+                          con = con, update.check.only = FALSE, ids_are_cultivars = FALSE)
2019-02-01 10:37:26 INFO   [query.trait.data] :
   ---------------------------------------------------------
2019-02-01 10:37:26 INFO   [query.trait.data] :
   stem_respiration_rate
Error in if (nrow(result) == 0) { : argument is of length zero
>

Stem respiration rate?

serbinsh · 2019-02-01T15:42:02Z

Perhaps this is it:

  } else if (trait == 'stem_respiration_rate') {
#########################  STEM RESPIRATION   ############################
    ## Apply Arrhenius scaling to convert stem respiration at measurement temp
    ## to that at 25 degC (ref temp).
    data <- arrhenius.scaling.traits(data = data, covariates = covariates, temp.covariates = c('leafT', 'airT'))

Covariate options are leafT or airT BUT many of these entries for stem respiration use stem temperature (stemT), e.g. https://modex.bnl.gov/bety/traits/41237

This can be addressed by added stemT as an option to this scale, and removing leafT (since I dont see how that is relevant) but raises my questions about covariates. we may need to make sure uploading includes the correct temp covariate where right now most of the time leafT is the only required. @dlebauer?

serbinsh · 2019-02-01T15:43:02Z

Looks like my problem is that there is 1 Quercus rubra value and that species is in my PFT but as mentioned the covariate scaling does not include stemT

serbinsh · 2019-02-01T15:52:59Z

Updating the function like this:

  } else if (trait == 'stem_respiration_rate') {
#########################  STEM RESPIRATION   ############################
    ## Apply Arrhenius scaling to convert stem respiration at measurement temp
    ## to that at 25 degC (ref temp).
    data <- arrhenius.scaling.traits(data = data, covariates = covariates, temp.covariates = c('stemT', 'airT','T'))

results in the expected results:

> data <- query.trait.data("stem_respiration_rate", PEcAn.utils::vecpaste(c("30","1014","1017","1130","1142","1170","1178","9303","9315")),
+                          con = con, update.check.only = FALSE, ids_are_cultivars = FALSE)
2019-02-01 10:48:35 INFO   [query.trait.data] :
   ---------------------------------------------------------
2019-02-01 10:48:35 INFO   [query.trait.data] :
   stem_respiration_rate
2019-02-01 10:48:35 INFO   [query.trait.data] :
   Median stem_respiration_rate : 11.72
2019-02-01 10:48:35 INFO   [query.trait.data] :
   ---------------------------------------------------------
> data
     id citation_id site_id treatment_id       name                date
1 41237         405     311         1268 High Light 1988-12-31 06:00:00
      time cultivar_id specie_id     mean statname      stat  n
1 00:00:00          NA      1170 11.72036       SE 0.4507832 NA
                  vname month       lon      lat control greenhouse
1 stem_respiration_rate    12 -89.42352 43.07746    TRUE       TRUE
  mean_unconverted stat_unconverted
1             10.4              0.4

infotroph · 2019-02-01T16:33:06Z

@serbinsh To make sure I'm following right: You're concluding that this is fully explained by the copy-paste error in stem resp scaling and doesn't need any changes to the if(nrow(result)==0) bit, yes?

serbinsh · 2019-02-01T16:34:13Z

LOL! yeah probably was a copy-paste error! Originally i thought it may be something else, but yeah, I think it was just the wrong covariates listed for stem respiration and my small PR seems to have fixed it

serbinsh · 2019-02-11T13:31:14Z

Closing since the fix was merged

serbinsh added Type: Bug Priority: 02 - Normal labels Feb 1, 2019

serbinsh assigned dlebauer, infotroph and serbinsh Feb 1, 2019

serbinsh added the Status: in progress label Feb 1, 2019

serbinsh mentioned this issue Feb 1, 2019

Minor fix to base/db/R/query.trait.data.R #2270

Merged

12 tasks

serbinsh closed this as completed Feb 11, 2019

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Bug in query.trait.data in relation to FRC_LC trait #2269

Bug in query.trait.data in relation to FRC_LC trait #2269

serbinsh commented Feb 1, 2019

serbinsh commented Feb 1, 2019

serbinsh commented Feb 1, 2019

serbinsh commented Feb 1, 2019

serbinsh commented Feb 1, 2019

serbinsh commented Feb 1, 2019

serbinsh commented Feb 1, 2019

infotroph commented Feb 1, 2019

serbinsh commented Feb 1, 2019

serbinsh commented Feb 11, 2019

Bug in query.trait.data in relation to FRC_LC trait #2269

Bug in query.trait.data in relation to FRC_LC trait #2269

Comments

serbinsh commented Feb 1, 2019

Bug Description

To Reproduce

serbinsh commented Feb 1, 2019

serbinsh commented Feb 1, 2019

serbinsh commented Feb 1, 2019

serbinsh commented Feb 1, 2019

serbinsh commented Feb 1, 2019

serbinsh commented Feb 1, 2019

infotroph commented Feb 1, 2019

serbinsh commented Feb 1, 2019

serbinsh commented Feb 11, 2019