How SAS Now Fully Supports Python for Data Science and AI

SAS, founded in 1976, announced full support for Python within its proprietary analytics software, enabling users to run Python code alongside SAS in the same environment.

Chief Technology Officer Bryan Harris said SAS embraces open‑source software to boost productivity, allowing analysts to choose either language without compromise.

SAS provides Python developers with code samples, including a PROC FCMP example that declares a Python object, defines and calls a Python function, and demonstrates a Black‑Scholes option pricing implementation.

/* A basic example of using PROC FCMP to execute a Python function */
proc fcmp;
declare object py(python);
submit into py;
def MyPythonFunction(arg1, arg2):
  "Output: ResultKey"
  Python_Out = arg1 * arg2
  return Python_Out
endsubmit;
rc = py.publish();
rc = py.call("MyPythonFunction", 5, 10);
SAS_Out = py.results["ResultKey"];
put SAS_Out=;
run;

def internal_black_scholes_call(stockPrice, strikePrice, timeRemaining, volatility, rate):
    import numpy
    from scipy import stats
    import math
    if ((strikePrice != 0) and (volatility != 0)):
        d1 = (math.log(stockPrice/strikePrice) + (rate + (volatility**2) /  2) * timeRemaining) / (volatility*math.sqrt(timeRemaining))
        d2 = d1 - (volatility * math.sqrt(timeRemaining))
        callPrice = (stockPrice * stats.norm.cdf(d1)) - (strikePrice * math.exp( (-rate) * timeRemaining) * stats.norm.cdf(d2))
    else:
        callPrice=0
    return (callPrice)

def black_scholes_call(stockPrice, strikePrice, timeRemaining, volatility, rate):
    "Output: optprice"
    import numpy
    from scipy import stats
    import math
    optPrice = internal_black_scholes_call(stockPrice, strikePrice, timeRemaining, volatility, rate)
    callPrice = float(optPrice)
    return (callPrice,)

Developers can build complete Python models in SAS Studio and integrate them with other SAS products such as Model Studio, Intelligent Decisioning, and Model Manager, supporting the full machine‑learning lifecycle.

SAS also announced a cloud‑native AI, analytics, and data‑management platform, Viya, with a local version and plans to add Kubernetes deployments in on‑premises environments, complementing its presence on Azure, AWS, Google Cloud, and Red Hat OpenShift.

The move mirrors similar strategies by companies like Neo4j, which also added native Python support to its data‑science toolkit.