Repository navigation
dbt Classes Instance Methods is_string, is_numeric, is_number, is_float don't seem to work #1380
Description
Describe the bug
The same issue with the methods persists in the Databricks adapter as was reported in the Spark adapter in this issue.
dbt Classes Instance Methods is_string, is_numeric, is_number, is_float don't seem to work
For our instance, col.is_string() does not works, and for one of our customers, the col.is_number() also does not work.
Steps To Reproduce
- Take any string field output from
get_columns_in_relationand print the result ofcol.is_string().
Expected behavior
- When
col.datatype == 'string'thencol.is_string()should evaluate to true.
Screenshots and log output
- I ran this print statement.
{{ print('Column: ' ~ col.name ~ ' | Data Type: ' ~ col.data_type ~ ' | Is Number: ' ~ col.is_number() ~ ' | Is String: ' ~ col.is_string()) }}
And the result shows it's not identifying correctly.
Column: name03 | Data Type: string | Is Number: False | Is String: False
System information
The output of dbt --version:
Core:
- installed: 1.11.6
- latest: 1.11.7 - Update available!
Plugins:
- databricks: 1.11.6 - Up to date!
The operating system you're using:
macOS
The output of python --version:
Python 3.13.9
Additional context
Add any other context about the problem here.
Activity
psaikaushik commentedon Apr 12, 2026
I can pick this up. The same issue was reported in dbt-spark (#658) and the root cause is the same — the
Columnbase class methods likeis_string()check against type names that don't include Spark/Databricks types.Looking at the code,
is_string()in the baseColumnclass likely checks for types liketext,character varying, etc. (PostgreSQL-style names), but Databricks returnsstringas the dtype. The fix is to override thestring_typeclassmethod (or the equivalent type classification method) in the DatabricksColumnimplementation to includestringin the set of recognized string types.Same pattern for
is_number()— Databricks usesint,bigint,double, etc. which may not all be in the base class's numeric type set.I'll look at the Column adapter class and send a PR.
fix: override is_string/is_number/is_float/is_integer on DatabricksCo…
psaikaushik commentedon Apr 12, 2026
PR up: #1390
Overrides
is_string(),is_number(),is_float(),is_integer(), andis_numeric()onDatabricksColumnwith the full set of Spark/Databricks type names. The baseColumnclass only knows PostgreSQL types, which is why these all returnedFalse.fix: override type classification methods for Databricks/Spark types (d…
fix: override is_string/is_number/is_float/is_integer on DatabricksCo…
test: add unit tests and changelog for type classification fix (datab…
test: add unit tests and changelog for type classification fix (datab…
test: add unit tests and changelog for type classification fix (datab…
github-actions commentedon Oct 25, 2026
This issue has been marked as Stale because it has been open for 180 days with no activity. If you would like the issue to remain open, please remove the stale label or comment on the issue.
fivetran-catfritz commentedon Oct 27, 2026
Yes this should still be fixed. No more stale please.