def _check_player_ids_need_update(self) -> bool:
"""
Helper function to check if player ids exist.
If it does, player ids need to be updated only if ``year`` does
not match THIS year.
"""
if self.player_ids_json_path.exists():
player_ids_loaded = utils.load_from_json(self.player_ids_json_path)
if player_ids_loaded.get("year") == self.year:
return False
return True
def test_write_to_and_load_from(tmp_path):
# Setup
tmp_file_path = tmp_path.joinpath("test.json")
json_data = {"test": "test", "test2": "test2"}
# Exercise
assert tmp_file_path.exists() is False # non-existent prior to write
utils.write_to_json(json_data, tmp_file_path)
result = utils.load_from_json(tmp_file_path)
# Verify
assert tmp_file_path.exists() is True
assert result == json_data
def test_load_from_json(tmp_path):
# Setup
tmp_file_path = tmp_path.joinpath("test.json")
json_data = {"test": "test", "test2": "test2"}
with open(tmp_file_path, "w") as written_file:
json.dump(json_data, written_file)
# Exercise
assert tmp_file_path.exists() is True # existent prior to load
result = utils.load_from_json(tmp_file_path)
# Verify
assert result == json_data
def test_load_stat_ids():
# Setup
file_loc = Path(__file__)
stat_ids_json_path = file_loc.parent.parent.joinpath(
"assets/stat_ids.json")
# Exercise
result = utils.load_from_json(stat_ids_json_path)
# Verify
assert len(result) == 91
for k, v in result.items():
assert int(k) == v["id"]
assert list(v.keys()) == ["id", "abbr", "name", "shortName"]
def test_load_player_ids():
# Setup
file_loc = Path(__file__)
stat_ids_json_path = file_loc.parent.parent.joinpath(
"assets/player_ids.json")
# Exercise
result = utils.load_from_json(stat_ids_json_path)
# Verify
assert "year" in result
for k, v in result.items():
if k == "year":
assert isinstance(v, int)
else:
assert list(v.keys()) == ["name", "position", "team", "injury"]
def create_player_pts_df(year: int,
week: int,
player_pts: Dict[str, Any],
savepath: Optional[Path] = None) -> pd.DataFrame:
"""
Create a DataFrame to house all player projected and actual points.
The ``player_pts`` argument can be ``projected_player_pts`` or
``actual_player_pts``.
Actual points will need to be pulled multiple times throughout as
more games are played/completed. Projected points should be pulled
once and only once.
"""
# Determine projected ('projectedStats') or actual points ('stats').
stats_type = _get_player_pts_stat_type(player_pts)
if stats_type == "projectedStats":
prefix = "PROJ_"
# Projected points should always be saved (pulled only once)
if savepath is None:
raise ValueError(
"When creating a projected player points dataframe, " +
"``savepath`` must be specified.")
# _get_player_pts_stat_type handles check and raises error if not
# projectedStats or stats
else:
prefix = "ACTUAL_"
index = []
points = []
for pid, player_pts_dict in player_pts.items():
points_dict = _unpack_player_pts(year, week, player_pts_dict)
index.append(pid)
points.append(points_dict)
player_pts_df = pd.DataFrame(points, index=index)
# Get definition of each point attribute
stat_ids_json_path = Path("assets/stat_ids.json")
stat_ids_dict = utils.load_from_json(stat_ids_json_path)
stat_defns = {
k: v["name"].replace(" ", "_")
for k, v in stat_ids_dict.items()
}
player_pts_df = player_pts_df.rename(columns=stat_defns)
player_pts_df = player_pts_df.add_prefix(prefix)
player_pts_df = player_pts_df.reset_index().rename(
columns={"index": "Player"})
# Get definition of each player team and name based on player id
player_ids_json_path = Path("assets/player_ids.json")
player_ids = utils.load_from_json(player_ids_json_path)
team = player_pts_df["Player"].apply(lambda x: player_ids[x]["team"])
player_pts_df.insert(1, "Team", team)
player_defns = {k: v["name"] for k, v in player_ids.items() if k != "year"}
name = player_pts_df["Player"].apply(lambda x: player_defns[x])
player_pts_df["Player"] = name
# Make pts col the third column for easy access
pts_col = player_pts_df.filter(regex=f"{prefix}pts")
pts_col_name = f"{prefix}pts"
player_pts_df = player_pts_df.drop(pts_col_name, axis=1)
player_pts_df.insert(2, pts_col_name, pts_col)
# Convert all col types to non-string as strings come from scrape
col_types = {}
for c in player_pts_df.columns:
if c in ("Player", "Team"):
col_types[c] = "object"
else:
col_types[c] = "float64"
player_pts_df = player_pts_df.astype(col_types)
player_pts_df = player_pts_df.fillna(0.0)
# Write projected players to csv so only done once
if stats_type == "projectedStats":
print(f"\tWriting projected player stats to {savepath}...")
player_pts_df.to_csv(savepath)
return player_pts_df