Stats not displayed for some CanESM2 climdices

[corviday]

CanESM2 rcp85 tnnETCCDI, tnxETCCDI, txxETCCDI, and txnWTCCDI return the following error message on the stats table:

selected[0] is undefined

Possibly an instance of this backend issue, which occurs when the backend attempts to aggregate data across files that use different units for the same variable. The four variables involved are among the handful known to have different units (degC vs. degrees_C), but I haven't verified this is the case.

[corviday]

SELECT DISTINCT 
  variable_aliases.variable_units, 
  models.model_short_name, 
  variable_aliases.variable_standard_name
FROM 
  ce_meta.ensembles, 
  ce_meta.ensemble_data_file_variables, 
  ce_meta.data_file_variables, 
  ce_meta.models, 
  ce_meta.data_files, 
  ce_meta.runs, 
  ce_meta.emissions, 
  ce_meta.variable_aliases
WHERE 
  ensembles.ensemble_id = ensemble_data_file_variables.ensemble_id AND
  ensemble_data_file_variables.data_file_variable_id = data_file_variables.data_file_variable_id AND
  data_file_variables.data_file_id = data_files.data_file_id AND
  data_file_variables.variable_alias_id = variable_aliases.variable_alias_id AND
  data_files.run_id = runs.run_id AND
  runs.model_id = models.model_id AND
  runs.emission_id = emissions.emission_id AND
  ensembles.ensemble_name LIKE 'ce_files' AND 
  models.model_short_name LIKE 'CanESM2' AND 
  variable_aliases.variable_standard_name LIKE 't%ETCCDI';

variable_units	model_short_name	Variable_standard_name
%	CanESM2	tn10pETCCDI
degC	CanESM2	tnnETCCDI
%	CanESM2	tc90ETCCDI
degC	CanESM2	tnxETCCDI
days	CanESM2	trETCCDI
degC	CanESM2	txnETCCDI
degC	CanESM2	txxETCCDI
%	CanESM2	tx10pETCCDI
%	CanESM2	tn90pETCCDI

Not a units issue.

[corviday]

Anecdotal observations have left me with the guess that the issue is the stats table is asking for monthly stats on annual-only datasets.

[rod-glover]

There's no variable tc90ETCCDI, but there is a tx90ETCCDI, which is not listed here. Assuming this is a typo.