🐝 update pandas to 2.2.2

apps/owidbot/data_diff.py

@@ -1,3 +1,4 @@
+import os
 import re
 import subprocess
 from typing import Tuple
@@ -89,7 +90,10 @@ def call_etl_diff(include: str) -> list[str]:
 
     print(" ".join(cmd))
 
-    result = subprocess.Popen(cmd, cwd=BASE_DIR, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
+    env = os.environ.copy()
+    env["PATH"] = os.path.expanduser("~/.cargo/bin") + ":" + env["PATH"]
+
+    result = subprocess.Popen(cmd, cwd=BASE_DIR, stdout=subprocess.PIPE, stderr=subprocess.PIPE, env=env)
     stdout, stderr = result.communicate()
 
     stdout = stdout.decode()

etl/steps/data/garden/artificial_intelligence/2023-07-25/cset.meta.yml

@@ -63,6 +63,7 @@ tables:
         unit: US dollars
         short_unit: $
         display:
+          name: << field >>
           numDecimalPlaces: 0
 
       investment_estimated:
@@ -88,6 +89,7 @@ tables:
         unit: US dollars
         short_unit: $
         display:
+          name: << field >>
           numDecimalPlaces: 0
 
       disclosed_investment_summary:
@@ -112,6 +114,7 @@ tables:
         unit: US dollars
         short_unit: $
         display:
+          name: << field >>
           numDecimalPlaces: 0
 
       estimated_investment_summary:
@@ -137,62 +140,71 @@ tables:
         unit: US dollars
         short_unit: $
         display:
+          name: << field >>
           numDecimalPlaces: 0
 
       num_patent_applications:
         title: Patent applications
         description: Yearly total number of patent applications.
         unit: applications
         display:
+          name: << field >>
           numDecimalPlaces: 0
 
       num_patent_granted:
         title: Patent applications granted
         description: Yearly total number of patents granted.
         unit: granted applications
         display:
+          name: << field >>
           numDecimalPlaces: 0
 
       num_patent_applications_summary:
         title: Total patent applications
         description: Total number of patent applications between 2010-2020.
         unit: applications
         display:
+          name: << field >>
           numDecimalPlaces: 0
 
       num_patent_granted_summary:
         title: Total patent applications granted
         description: Total number of patents granted between 2010-2020.
         unit: granted applications
         display:
+          name: << field >>
           numDecimalPlaces: 0
 
       num_articles:
         title: Number of articles
         description: Yearly number of articles published in AI fields.
         unit: articles
         display:
+          name: << field >>
           numDecimalPlaces: 0
 
       num_citations:
         title: Number of citations
         description: Yearly number of citations in the AI field.
         unit: citations
         display:
+          name: << field >>
           numDecimalPlaces: 0
 
       num_articles_summary:
         title: Total number of articles
         description: Yearly total number of articles in the AI field between 2010-2022
         unit: articles
         display:
+          name: << field >>
           numDecimalPlaces: 0
 
       num_citations_summary:
         title: Total number of citations
         description: The total number of citations in the AI field between 2010-2022
         unit: citations
         display:
+          name: << field >>
           numDecimalPlaces: 0
 
       proportion_patents_granted:
@@ -201,6 +213,7 @@ tables:
         unit: '%'
         short_unit: '%'
         display:
+          name: << field >>
           numDecimalPlaces: 0
 
       citations_per_article:
@@ -209,25 +222,29 @@ tables:
         unit: 'citations per article'
         short_unit: ''
         display:
+          name: << field >>
           numDecimalPlaces: 0
 
       num_patent_applications_per_mil:
         title: Patent applications per 1 million people
         description: Yearly total number of patent applications per 1 million people.
         unit: applications
         display:
+          name: << field >>
           numDecimalPlaces: 0
 
       num_patent_granted_per_mil:
         title: Patent applications granted per 1 million people
         description: Yearly total number of patents granted per 1 million people.
         unit: granted applications
         display:
+          name: << field >>
           numDecimalPlaces: 0
 
       num_articles_per_mil:
         title: Number of articles per 1 million people
         description: Yearly number of articles published in AI fields per 1 million people
         unit: articles
         display:
-          numDecimalPlaces: 0
+          name: << field >>
+          numDecimalPlaces: 0

etl/steps/data/grapher/artificial_intelligence/2023-07-25/cset.py

@@ -4,7 +4,6 @@
 
 from owid.catalog import Dataset
 
-from etl import grapher_helpers as gh
 from etl.helpers import PathFinder, create_dataset, grapher_checks
 
 # Get paths and naming conventions for current step.
@@ -23,20 +22,11 @@ def run(dest_dir: str) -> None:
     tb = tb.reset_index()
     tb.set_index(["country", "year", "field"], inplace=True)
 
-    # Expand dimensions into columns.
-    expanded_tb = gh.expand_dimensions(tb)
-
-    # Set display name to its `field` for each column
-    for col in expanded_tb.columns:
-        dim_filters = expanded_tb[col].metadata.additional_info["dimensions"]["filters"]
-        assert len(dim_filters) == 1
-        expanded_tb[col].metadata.display["name"] = dim_filters[0]["value"]
-
     #
     # Save outputs.
     #
     # Create a new grapher dataset with the same metadata as the garden dataset
-    ds_grapher = create_dataset(dest_dir, tables=[expanded_tb], default_metadata=ds_garden.metadata)
+    ds_grapher = create_dataset(dest_dir, tables=[tb], default_metadata=ds_garden.metadata)
 
     #
     # Checks.

lib/catalog/pyproject.toml

@@ -8,7 +8,7 @@ authors = [
 license = "MIT"
 requires-python = ">=3.9, <4.0"
 dependencies = [
-    "pandas==2.2.1",
+    "pandas>=2.2.1",
     "jsonschema>=3.2.0",
     "pyarrow>=10.0.1",
     "ipdb>=0.13.9",

lib/datautils/pyproject.toml

@@ -8,7 +8,7 @@ authors = [
 license = "MIT"
 requires-python = ">=3.9, <4.0"
 dependencies = [
-    "pandas==2.2.1",
+    "pandas>=2.2.1",
     "boto3>=1.21.16",
     "structlog>=21.5.0",
     "colorama>=0.4.4",

lib/repack/pyproject.toml

@@ -9,7 +9,7 @@ license = "MIT"
 requires-python = ">=3.9"
 dependencies = [
     "numpy>=1.24.0",
-    "pandas==2.2.1"
+    "pandas>=2.2.1"
 ]
 
 [tool.uv]