GCAM-Europe double slash queries work for GCAM-core regions but not for GCAM-Europe regions

[klau506]

rgcam::runQuery() fails for GCAM-Europe regions but works for GCAM-core regions

When running queries that require double slash through rgcam, rgcam::runQuery(conn, co2_query, scenario, reg), the function fails whenever reg corresponds to a GCAM-Europe region, while it runs correctly for GCAM-core regions.

For example, if we introduce a double slash between subsector and technology:

co2_query <- "<emissionsQueryBuilder title=\"CO2 emissions by tech (nested subsector) (excluding resource production)\">
  <axis1 name=\"technology\">technology</axis1>
  <axis2 name=\"Year\" optional=\"true\">emissions</axis2>
  <xPath buildList=\"true\" dataName=\"emissions\" group=\"false\" sumAll=\"false\" optional=\"true\">
    *[@type = 'sector' ]/*[@type='subsector']//*[@type='technology')]//CO2/emissions/node()
  </xPath>
  <comments/>
</emissionsQueryBuilder>"

---

The same query runs successfully in Model Interface, but fails in the rgcam package when querying GCAM-Europe regions.

Inspecting the output using the MI, we observed two irregularities when running the query with and without the double slash between subsector and technology:

1. Different values for some sector–subsector–technology entries

Example output differences:

scenario	region	sector	subsector	technology	1990	2005	2010	2015	2020	2025	2030	2035	2040	2045	2050	2065	2080	2095	Units
Ref (query with single slash)	Spain	alumina	gas	gas	0.00662193	0.00399082	0.0529131	0.0772193	0.0889154	0.09128569	0.08693893	0.07746966	0.0675153	0.06089516	0.056721945	0	0	0	MTC
Ref (query with double slash)	Spain	alumina	gas	gas	0.01324386	0.00798164	0.1058262	0.1544386	0.1778308	0.18257138	0.17387786	0.15493932	0.1350306	0.12179032	0.11344389	0	0	0	MTC

2. New EU electricity-grid-related sectors missing unless // is used in XPath

The GCAM-Europe version introduces new electricity-grid sectors, which are base load generation, intermediate generation, peak generation, and subpeak generation. These only appear when XPath includes a double slash.

---

The exact error message when running:

rgcam::runQuery(conn, co2_query, scenario, "Spain")

is:

About to perform query: collection()/scenario
About to perform query: collection()/scenario[ (@name='EU_NECP_LTT-Rev1' and @date='2025-10-11T09:44:43+01:00') ]/world/region[ (@name='Spain') ]/*[@type = 'sector' ]/*[@type='subsector']//*[@type='technology')]// CO2/emissions/node()
In Function: 1765368225619
java.lang.ArrayIndexOutOfBoundsException: 8
    at ModelInterface.ModelGUI2.xmldb.RunMIQuery.buildTable(RunMIQuery.java:188)
    ...
Stopped at ...: java.lang.ArrayIndexOutOfBoundsException: 8
Rows: 0 Columns: 0

while, for instance running:

rgcam::runQuery(conn, co2_query, scenario, "USA")

works well.

---

@pralitp, do you have any idea why this query:

works correctly in MI, but fails when executed through rgcam, specifically for GCAM-Europe regions?

We though it could be related to:

• differences in sector hierarchy introduced by the EU grid disaggregation?

• an XPath expression that MI tolerates but rgcam/BaseX does not?

• assumptions in RunMIQuery.buildTable() that break with deeper or duplicated structures?

Any guidance would be very helpful! Thank you in advance!!