Sportsipy API request - api

Question

I need to use the sportsipy API to get the schedule for all teams in a dataframe. This is what I have:
from sportsreference.nba.schedule import Schedule
league = ['MIL','CHO','LAL','LAC','SAC','ATL','MIA','DAL','POR',
'HOU','NOP','PHO','WAS','MEM','BOS','DEN','TOR','SAS',
'PHI','BRK','UTA','IND','OKC','ORL','MIN','DET',
'NYK','CLE','CHI','GSW']
for i in league:
mil2019 = Schedule( i , year = '2020')
mil2019.dataframe_extended
The error i get is:
TypeError: unsupported operand type(s) for -: 'NoneType' and 'NoneType'

Answer 1

As mentioned above in the comment above, I believe your import is wrong. Using package sportsipy 0.6.0 and following the docs: https://sportsipy.readthedocs.io/en/stable/ I was able to achieve your desired result using following code:
from sportsipy.nba.schedule import Schedule
# MIL removed from league list as it is used to initiate league_schedule
league = ['CHO','LAL','LAC','SAC','ATL','MIA','DAL','POR',
'HOU','NOP','PHO','WAS','MEM','BOS','DEN','TOR','SAS',
'PHI','BRK','UTA','IND','OKC','ORL','MIN','DET',
'NYK','CLE','CHI','GSW']
league_schedule = Schedule('MIL', year="2020").dataframe
for team in league:
league_schedule = league_schedule.append(Schedule(team , year="2020").dataframe)
(Resulting dataframe has dimensions: [2286 rows x 15 columns])
Same should work with dataframe_extended, but it takes rather long time to get all that data. Maybe double check if you need all of it.
In case I am wrong and package you want to use in your question is correct please add additional info to your question, such as where can we get that package.

Answer 2

It appears you are using the module from pip install sportsreference from here which is on v0.5.2 in which case that is a valid import, even though you mentioned you're using sportsipy which caused some confusion for others. The latest version has refactored the package name to sportsipy.
If it wasn't a valid import, it would be throwing an import error on the very first line, so I'm not sure why folks are getting hung up on that.
You really should include the entire Python traceback, not just the final message, so we can determine exactly where in your code and the module's source code this exception is being raised. Also include the specific version of the library you're using, e.g. from pip freeze.
My initial thought is one of the requests somewhere for one of these teams is returning something unexpected and the library is not handling it properly, but without the full traceback that's just a theory.
It's probably a bug in v0.5.2 of sportsipy. I would try using the latest version from git and see if you can reproduce the error. Something, somewhere isn't validating things are what it expects before trying to do things with it. If I had the full traceback, I could tell you exactly what.
You could try catching the TypeError and passing on it, to see if skipping it allows everything else to continue working, but without knowing exactly where the error is coming from it's hard to say for sure at this point.
for i in league:
try:
mil2019 = Schedule( i , year = '2020')
mil2019.dataframe_extended
except TypeError:
pass
This won't fix the problem, it's actually hiding it, but if it's just one record from one game that is returning something unexpected this would at least let you get the rest of the results, possibly. It's also possible the issue would create other problems later, depending on exactly what it is. Again, this is where the whole traceback would have been helpful.
I will say that trying your code for just one team works for me. For example:
from sportsreference.nba.schedule import Schedule
mil2019 = Schedule("MIL", year="2020")
print(mil2019.dataframe_extended.head(10))
Returns this:
away_assist_percentage ... winning_name
201910240HOU 67.4 ... Milwaukee Bucks
201910260MIL 71.7 ... Miami Heat
201910280MIL 42.2 ... Cleveland Cavaliers
201910300BOS 55.3 ... Milwaukee Bucks
201911010ORL 51.1 ... Milwaukee Bucks
201911020MIL 70.6 ... Toronto Raptors
201911040MIN 48.0 ... Milwaukee Bucks
201911060LAC 42.9 ... Milwaukee Bucks
201911080UTA 41.2 ... Milwaukee Bucks
201911100OKC 57.4 ... Milwaukee Bucks
[10 rows x 82 columns]
It takes forever just to get the games from one team. The library is not passing around an existing requests.Session() when calling PyQuery (even though PyQuery supports a session kwarg), so every request for every box score is renegotiating a fresh TCP connection which is absurd, but I digress:
DEBUG:urllib3.connectionpool:Starting new HTTP connection (1): www.basketball-reference.com:80
DEBUG:urllib3.connectionpool:http://www.basketball-reference.com:80 "GET /teams/MIL/2020_games.html HTTP/1.1" 301 183
DEBUG:urllib3.connectionpool:Starting new HTTPS connection (1): www.basketball-reference.com:443
DEBUG:urllib3.connectionpool:https://www.basketball-reference.com:443 "GET /teams/MIL/2020_games.html HTTP/1.1" 200 34571
DEBUG:urllib3.connectionpool:Starting new HTTPS connection (1): www.basketball-reference.com:443
DEBUG:urllib3.connectionpool:https://www.basketball-reference.com:443 "GET /boxscores/201910240HOU.html HTTP/1.1" 200 46549
DEBUG:urllib3.connectionpool:Starting new HTTPS connection (1): www.basketball-reference.com:443
I would add some debugging to your code to establish which team your code is working on when this exception is raised. Try first with one team like I did and confirm it generally works, then iterate through the list of teams with logging enabled like:
import logging
from sportsreference.nba.schedule import Schedule
logging.basicConfig(level=logging.DEBUG)
league = ['CHO', 'LAL', 'LAC', 'SAC', 'ATL', 'MIA', 'DAL', 'POR',
'HOU', 'NOP', 'PHO', 'WAS', 'MEM', 'BOS', 'DEN', 'TOR', 'SAS',
'PHI', 'BRK', 'UTA', 'IND', 'OKC', 'ORL', 'MIN', 'DET',
'NYK', 'CLE', 'CHI', 'GSW']
for i in league:
logging.info("Working on league: %s", i)
mil2019 = Schedule(i, year="2020")
print(mil2019.dataframe_extended)
This way you will know specifically what league and which specific request is responsible for the issue and that would help you determine what the root cause is.