Moving older scripts into holding cell

.gitignore

@@ -132,3 +132,4 @@ dmypy.json
 *.db
 .idea
 theshow/
+.vscode/

creation_helpers.py

@@ -2,7 +2,7 @@ import csv
 import datetime
 import logging
 import math
-from decimal import Decimal
+from decimal import ROUND_HALF_EVEN, Decimal
 
 import pandas as pd
 import pybaseball as pb
@@ -860,10 +860,38 @@ def defense_rg(all_pos: list) -> list:
 
 
 def sanitize_chance_output(total_chances, min_chances=1.0, rounding=0.05):
+    if total_chances < min_chances:
+        logging.debug(f'sanitize: {total_chances} is less than min_chances ({min_chances}); returning 0')
+        return 0
+    
+    rounded_decimal = round(Decimal(str(total_chances)) / Decimal(str(rounding))) * Decimal(str(rounding))
+
+    exact_chances = [
+        Decimal('1.05'), Decimal('1.1'), Decimal('1.2'), Decimal('1.25'), Decimal('1.3'), Decimal('1.35'),
+        Decimal('1.4'), Decimal('1.5'), Decimal('1.6'), Decimal('1.65'), Decimal('1.7'), Decimal('1.75'),
+        Decimal('1.8'), Decimal('1.9'), Decimal('1.95'), Decimal('2.1'), Decimal('2.2'), Decimal('2.25'),
+        Decimal('2.4'), Decimal('2.5'), Decimal('2.55'), Decimal('2.6'), Decimal('2.7'), Decimal('2.75'),
+        Decimal('2.8'), Decimal('2.85'), Decimal('3.2'), Decimal('3.25'), Decimal('3.3'), Decimal('3.4'),
+        Decimal('3.5'), Decimal('3.6'), Decimal('3.75'), Decimal('3.8'), Decimal('3.9'), Decimal('4.2'),
+        Decimal('4.25'), Decimal('4.5'), Decimal('4.75'), Decimal('4.8'), Decimal('5.1'), Decimal('5.4'),
+        Decimal('5.7')
+    ]
+
+    if rounded_decimal > exact_chances[-1]:
+        return float(rounded_decimal)
+    
+    for x in exact_chances:
+        if rounded_decimal <= x:
+            return float(x)
+    
+    raise ArithmeticError(f'Attempt to sanitize {total_chances} rounded to {rounded_decimal} and could not be matched to an exact result')
+
+
+def legacy_sanitize_chance_output(total_chances: float, min_chances: float = 1.0, rounding: float = 0.05):
     # r_val = mround(total_chances) if total_chances >= min_chances else 0
     r_val = Decimal(total_chances) if total_chances >= min_chances else Decimal(0)
     logging.debug(f'r_val: {r_val}')
-    rounded_val = Decimal(float(math.floor(r_val / Decimal(rounding)) * Decimal(rounding))).quantize(Decimal("0.05"))
+    rounded_val = Decimal(float(math.floor(r_val / Decimal(rounding)) * Decimal(rounding))).quantize(Decimal("0.05"), ROUND_HALF_EVEN)
     if math.floor(rounded_val) == rounded_val:
         return rounded_val
 
@@ -885,6 +913,7 @@ def sanitize_chance_output(total_chances, min_chances=1.0, rounding=0.05):
             return x
 
 
+
 def mlbteam_and_franchise(mlbam_playerid):
     api_url = f'https://statsapi.mlb.com/api/v1/people/{mlbam_playerid}?hydrate=currentTeam'
     logging.info(f'Calling {api_url}')

db_calls.py

@@ -7,7 +7,7 @@ from typing import Literal, Optional
 AUTH_TOKEN = {'Authorization': f'Bearer Tp3aO3jhYve5NJF1IqOmJTmk'}
 DB_URL = 'https://pd.manticorum.com/api'
 master_debug = True
-alt_database = False
+alt_database = 'dev'
 
 if alt_database == 'dev':
     DB_URL = 'https://pddev.manticorum.com/api'

pybaseball_doodling.py

@@ -0,0 +1,97 @@
+import asyncio
+import requests
+import sys
+import time
+
+from bs4 import BeautifulSoup
+import pandas as pd
+import pybaseball as pb
+
+from app.creation_helpers import get_pitching_peripherals
+
+
+def pull_stats(start_date_string: str, end_date_string: str):
+    print(f'Pulling stats from {start_date_string} to {end_date_string}')
+    data = pb.batting_stats_range(start_date_string, end_date_string)
+    print(f'Got it!')
+    return data
+
+
+def batting_game_log_table(bbref_id: str, season: int = 1998):
+    page_url = f'https://www.baseball-reference.com/players/gl.fcgi?id={bbref_id}&t=b&year={season}'
+    soup = BeautifulSoup(requests.get(page_url).text, 'html.parser')
+    time.sleep(3)
+    table = soup.find('table', {'id': 'batting_gamelogs'})
+    tbody = table.find('tbody')
+    headers = []
+    data = []
+    indeces = []
+
+    for row in tbody.find_all('tr'):
+        row_data = []
+        col_names = []
+        for cell in row.find_all('td'):
+            row_data.append(cell.text)
+            if len(headers) == 0:
+                col_names.append(cell['data-stat'])
+
+        if len(row_data) > 0:
+            data.append(row_data)
+            indeces.append(row_data[0])
+            if len(headers) == 0:
+                headers.extend(col_names)
+    
+    player_frame = pd.DataFrame(data, index=indeces, columns=headers)
+    return player_frame 
+
+
+def batting_splits_table(bbref_id: str, season: int = 1998):
+    page_url = f'https://www.baseball-reference.com/players/split.fcgi?id={bbref_id}&year={season}&t=b'
+    soup = BeautifulSoup(requests.get(page_url).text, 'lxml')
+    time.sleep(3)
+    table = soup.find('table', {'id': 'plato'})
+    tbody = table.find('tbody')
+    headers = []
+    data = []
+    indeces = []
+
+    for row in tbody.findAll('tr'):
+        row_data = []
+        col_names = []
+        for cell in row.find_all('td'):
+            row_data.append(cell.text)
+            if len(headers) == 0:
+                col_names.append(cell['data-stat'])
+        
+        if len(row_data) > 0:
+            data.append(row_data)
+            indeces.append(row_data[0])
+            if len(headers) == 0:
+                headers.extend(col_names)
+    
+    bsplit_frame = pd.DataFrame(data, index=indeces, columns=headers)
+    return bsplit_frame
+
+
+async def main(args):
+    # print('Fetching peripherals')
+    # peripherals = get_pitching_peripherals(2021)
+    # print('Got them!')
+    # print(f'There are {len(peripherals)} records')
+
+    # month_1 = pull_stats('1997-03-01', '1997-04-30')
+    # print(f'Received {len(month_1)} records.')
+
+    # print(f'Grabbing game logs...')
+    # player_frame = batting_game_log_table('jeterde01')
+    # print(f'Received {len(player_frame)} records')
+
+    print(f'Grabbing game logs...')
+    player_frame = batting_splits_table('jeterde01')
+    print(f'Received {len(player_frame)} records')
+
+    pass
+
+
+if __name__ == '__main__':
+    asyncio.run(main(sys.argv[1:]))

pytest.ini

@@ -0,0 +1,4 @@
+[pytest]
+asyncio_mode = auto
+filterwarnings =
+    ignore::DeprecationWarning

requirements.txt

@@ -0,0 +1,51 @@
+aiohappyeyeballs==2.4.3
+aiohttp==3.10.10
+aiosignal==1.3.1
+annotated-types==0.7.0
+attrs==24.2.0
+beautifulsoup4==4.12.3
+bs4==0.0.2
+certifi==2024.8.30
+cffi==1.17.1
+charset-normalizer==3.4.0
+contourpy==1.3.0
+cryptography==43.0.1
+cycler==0.12.1
+Deprecated==1.2.14
+fonttools==4.54.1
+frozenlist==1.4.1
+idna==3.10
+iniconfig==2.0.0
+kiwisolver==1.4.7
+lxml==5.3.0
+matplotlib==3.9.2
+multidict==6.1.0
+numpy==2.1.2
+packaging==24.1
+pandas==2.2.3
+pillow==11.0.0
+pluggy==1.5.0
+propcache==0.2.0
+pyarrow==17.0.0
+pybaseball==2.2.7
+pycparser==2.22
+pydantic==2.9.2
+pydantic_core==2.23.4
+PyGithub==2.4.0
+PyJWT==2.9.0
+PyNaCl==1.5.0
+pyparsing==3.2.0
+pytest==8.3.3
+pytest-asyncio==0.24.0
+python-dateutil==2.9.0.post0
+pytz==2024.2
+requests==2.32.3
+scipy==1.14.1
+six==1.16.0
+soupsieve==2.6
+tqdm==4.66.5
+typing_extensions==4.12.2
+tzdata==2024.2
+urllib3==2.2.3
+wrapt==1.16.0
+yarl==1.15.4

tests/test_batter_calcs.py

@@ -0,0 +1,68 @@
+from decimal import ROUND_HALF_EVEN, Decimal
+import math
+
+from batters.calcs_batter import bp_singles, wh_singles, sanitize_chance_output
+from creation_helpers import sanitize_chance_output, mround
+
+
+def test_sanitize():
+    # def my_round_decimal(i: float):
+    #     return Decimal(i).quantize(Decimal('0.05'), ROUND_HALF_EVEN)
+
+    # def my_round(num: float, to: float = 0.05):
+    #     num, to = Decimal(str(num)), Decimal(str(to))
+    #     return float(round(num / to) * to)
+    
+    # assert my_round(6) == 6
+    # assert my_round(5.96) == 5.95
+    # assert my_round(5.84) == 5.85
+    # assert my_round(3.123) == 3.1
+    # assert math.floor(my_round(6)) == 6
+    # assert math.floor(my_round(5.96)) == 5
+
+    assert sanitize_chance_output(6) == 6.0
+    assert sanitize_chance_output(1.21) == 1.2
+    assert sanitize_chance_output(4.77) == 4.75
+    assert sanitize_chance_output(4.78) == 4.8
+
+    # step_1 = Decimal(6) / Decimal(0.05)
+    # step_1_5 = round(step_1, )
+    # step_2 = round(step_1)
+    # step_3 = float(step_2 * Decimal(0.05))
+    # step_4 = Decimal(step_3)
+
+    # assert Decimal(6) == Decimal(6).quantize(Decimal('0.05'), ROUND_HALF_EVEN)
+
+    # assert round(step_1) == 120
+    # # assert step_1 == 120
+    # assert step_1_5 == 120
+    # assert step_2 == 120
+    # assert step_3 == 6.0
+    # assert step_4 == Decimal('6.0')
+    
+    # rounded_val = step_4.quantize(Decimal("0.05"), ROUND_HALF_EVEN)
+    # assert rounded_val == 6
+    # assert sanitize_chance_output(6) == 6
+
+
+def test_mround():
+    assert mround(6) == 6.0
+    assert mround(1.21) == 1.2
+    assert mround(4.77) == 4.75
+    assert mround(4.78) == 4.8
+
+
+def test_decimals():
+    assert Decimal(8) == 8
+
+
+def test_bp_singles():
+    assert bp_singles(23) == 5
+    assert bp_singles(6) == 5
+    assert bp_singles(3) == 0
+
+
+def test_wh_singles():
+    assert wh_singles(11, .36) == 3.75
+    assert wh_singles(12, .45) == Decimal('7.95')
+

tests/test_helpers.py

@@ -0,0 +1,7 @@
+from creation_helpers import pd_positions_df
+
+
+def test_positions_df():
+    cardset_19_pos = pd_positions_df(19)
+
+    assert True == True