Merge pull request #721 from ElectionDataAnalysis/feature-version-2.0

Feature version 2.0

Author: sfsinger19103

docs/About_the_Code.md

@@ -18,6 +18,25 @@ The information in this directory determines the structure of the database creat
 
 ### Conventions
  * Some jurisdiction names (e.g., "District of Columbia") contain spaces, while it is inconvenient to use spaces in directory and file names. So we distinguish between a jurisdiction's "true name" and its "system name", which replaces all spaces by hyphens (e.g., "District-of-Columbia").
+
+
+### Testing
+There are `pytest` routines to test the dataloading, analyzer and jurisdiction-prepping functions. The data required to test the latter two is in the [`000_data_for_pytest`](../tests/000_data_for_pytest) folder. These depend on a file `tests/run_time.ini`. Note that the run_time.ini file is *not* part of the repository, as it contains database connection information. You will need to create it yourself. Or you can point to a different parameter file with the custom option `--param_file` for  pytest.
+
+The [dataloader test](../tests/dataloading_tests/test_dataloading.py) depends on information outside the [tests folder] (../tests):
+ - election results files in the results directory specified in by the `results_dir` parameter in `dataloading_tests/run_time.ini`. 
+    - if the results directory does not exist, the test will create it and pull files from [`https://github.com/ElectionDataAnalysis/TestingData.git`](https://github.com/ElectionDataAnalysis/TestingData.git). You can specify a different url with the custom pytest option `--test_data_url`
+ - reference files in the [`reference_results` folder](../src/reference_results). By convention, these are tabs-separated and named for the jurisdiction, e.g., `Virgina.tsv` or `American-Samoa.tsv`. Note the hyphens. If there are no reference results for a given election-jurisdiction pair, the test will fail. The reference files must have columns `Jurisdiction,Election,Contest,ReportingUnit,VoteType,Count`. 
+
+Note that the `analyzer_tests` and `dataloader_tests` directories each have a `conftest.py` file. This may cause a problem if you try to run them simultaneously via `pytest` from the `test` directory. Running them separately works:
+```
+tests % pytest analyzer_tests
+tests % pytest dataloader_tests
+```
+
+
+
+
 
 
 

docs/User_Guide.md

@@ -1,6 +1,6 @@
 matplotlib==3.1.2
 numpy==1.18.1
-pandas==1.1.0
+pandas==1.3.2
 psycopg2==2.8.6
 scipy==1.6.3
 SQLAlchemy==1.3.12

requirements.txt

@@ -6,7 +6,7 @@
     package_dir={"": "src"},
     packages=find_packages(where="src"),
     include_package_data=True,
-    url="https://github.com/ElectionDataAnalysis/election_data_analysis",
+    url="https://github.com/ElectionDataAnalysis/electiondata",
     author="Stephanie Frank Singer, et al.",
     author_email="sfsinger@campaignscientific.com",
     install_requires=["sqlalchemy", "pandas"],