Merge pull request #235 from openscm/dependabot/pip/pandas-gte-1.1-an…

…d-lt-3

Update pandas requirement from <2,>=1.1 to >=1.1,<3

.github/workflows/ci.yaml

@@ -106,7 +106,6 @@ jobs:
           poetry run pytest -r a -v src tests
 
   test-no-plotting:
-    # seaborn imports scipy which can make it available without being explicit.
     # This test makes sure that everything works even if the plotting
     # dependencies aren't installed.
     runs-on: ubuntu-latest
@@ -121,7 +120,7 @@ jobs:
         with:
           python-version: "${{ matrix.python-version }}"
           venv-id: "tests-${{ runner.os }}"
-          poetry-dependency-install-flags: "--all-extras --only 'tests'"
+          poetry-dependency-install-flags: "--extras optional --with tests"
       - name: Test with pytest
         run: |
           poetry run pytest -r a -v src tests

changelog/235.improvement.md

@@ -0,0 +1,3 @@
+Added support for pandas>=2
+
+Requirement now set to 'pandas>=1.1'

changelog/235.trivial.md

@@ -0,0 +1,3 @@
+Fixed broken CI of no plotting test
+
+Fixes a regression when translating the old no plotting test to the new poetry CI setup in [#260](https://github.com/openscm/scmdata/pull/260)

docs/source/changelog.md

@@ -19,8 +19,7 @@ of rst and use slightly different categories.
 
 <!-- towncrier release notes start -->
 
-v0.15.2
--------
+## v0.15.2
 
 - ([#257](https://github.com/openscm/scmdata/pull/257)) Updated to support the latest version of {mod}`notebook`
 - ([#252](https://github.com/openscm/scmdata/pull/252)) Add ``py.typed`` file do downstream packages can use

pyproject.toml

@@ -32,7 +32,7 @@ cftime = ">=1.5"
 numpy = "*"
 openscm-units = "*"
 packaging = "*"
-pandas = ">=1.1, <2"
+pandas = ">=1.1"
 pint = "<0.20"
 pint-pandas = "*"
 python-dateutil = "*"

tests/unit/test_pyam_compat.py

@@ -1,6 +1,7 @@
 import re
 from unittest import mock
 
+import pandas as pd
 import pytest
 
 from scmdata.pyam_compat import LongDatetimeIamDataFrame
@@ -13,9 +14,18 @@ def test_to_int_value_error(test_iam_df):
     postion = 4
     idf.loc[postion, "time"] = bad_val
 
-    error_msg = re.escape(
-        f"Unknown string format: {bad_val} present at position {postion}"
-    )
+    if pd.__version__.startswith("1"):
+        error_msg = re.escape(
+            f"Unknown string format: {bad_val} present at position {postion}"
+        )
+
+    else:
+        error_msg = re.escape(
+            f'time data "{bad_val}" '
+            "doesn't match format "
+            '"%Y/%m/%d", at position 4. You might want to try:'
+        )
+
     with pytest.raises(ValueError, match=error_msg):
         LongDatetimeIamDataFrame(idf)