y = x - 1
				i = 2
			}
		}
		switch i {
		case 0:
			z := y * y
			p1 := _lgamA[0] + z*(_lgamA[2]+z*(_lgamA[4]+z*(_lgamA[6]+z*(_lgamA[8]+z*_lgamA[10]))))
			p2 := z * (_lgamA[1] + z*(+_lgamA[3]+z*(_lgamA[5]+z*(_lgamA[7]+z*(_lgamA[9]+z*_lgamA[11])))))
			p := y*p1 + p2
			lgamma += (p - 0.5*y)
		case 1:
			z := y * y
			w := z * y

        y = Conj(scope_, x);
        break;
      case COMPLEX:
        y = Complex(scope_, x, x);
        break;
      case ANGLE:
        y = Angle(scope_, x);
        break;
      case LGAMMA:
        y = Lgamma(scope_, x);
        break;
      case ERF:
        y = Erf(scope_, x);
        break;
      case ERFINV:
        y = Erfinv(scope_, x);
        break;
      case NDTRI:

	} else {
		y1 = Pow(x, x-0.5) / y1
	}
	return y1, SqrtTwoPi * w * y2
}

// Gamma returns the Gamma function of x.
//
// Special cases are:
//
//	Gamma(+Inf) = +Inf
//	Gamma(+0) = +Inf
//	Gamma(-0) = -Inf
//	Gamma(x) = NaN for integer x < 0
//	Gamma(-Inf) = NaN
//	Gamma(NaN) = NaN
func Gamma(x float64) float64 {
	const Euler = 0.57721566490153286060651209008240243104215933593992 // A001620

    // Compute the reflected value, used when x < 0.5:
    //
    //   lgamma(x) = log(pi) - lgamma(1-x) - log(abs(sin(pi * x))).
    //
    // (The abs is because lgamma is the log of the absolute value of the gamma
    // function.)
    //
    // We have to be careful when computing the final term above. gamma(x) goes
    // to +/-inf at every integer x < 0, and this is controlled by the

	for i := 0; i < len(vf); i++ {
		if f, s := Lgamma(vf[i]); !close(lgamma[i].f, f) || lgamma[i].i != s {
			t.Errorf("Lgamma(%g) = %g, %d, want %g, %d", vf[i], f, s, lgamma[i].f, lgamma[i].i)
		}
	}
	for i := 0; i < len(vflgammaSC); i++ {
		if f, s := Lgamma(vflgammaSC[i]); !alike(lgammaSC[i].f, f) || lgammaSC[i].i != s {
			t.Errorf("Lgamma(%g) = %g, %d, want %g, %d", vflgammaSC[i], f, s, lgammaSC[i].f, lgammaSC[i].i)
		}
	}
}

  %0 = "tf.Round"(%arg0) : (tensor<*xf32>) -> tensor<*xf32>
  func.return %0 : tensor<*xf32>
}

// CHECK-LABEL: func @lgamma
func.func @lgamma(%arg0: tensor<4xf32>) -> tensor<4xf32> {
  // The lowering for lgamma is complicated, which makes it awkward to write a
  // complete test for it here. Instead we test that Lgamma is at least being
  // lowered here and rely on UnaryOpsTest.testFloatOps and other TensorFlow

  Scope grad_scope = scope.WithControlDependencies(grad);
  auto x = ConjugateHelper(grad_scope, op.input(0));
  auto dx = Mul(grad_scope, grad, Digamma(grad_scope, x));
  grad_outputs->push_back(dx);
  return grad_scope.status();
}
REGISTER_GRADIENT_OP("Lgamma", LgammaGrad);

Status MinOrMaxGrad(const Scope& scope, const Operation& op,
                    const std::vector<Output>& grad_inputs,

            "Rsqrt", "Sigmoid", "Sign", "Sinh", "Softplus", "Softsign", "Sqrt",
            "Square", "Tan", "Tanh", "Real", "Imag", "Erf", "Erfc", "Erfinv",
            "Lgamma", "Digamma",
            // Binary
            "Add", "AddV2", "Sub", "Mul", "Div", "Atan2", "Complex", "DivNoNan",
            "MulNoNan", "FloorDiv", "Xlogy", "Xlog1py", "Xdivy", "FloorMod",

Alexander Zielenski <******@****.***> 1697824555 -0700

		}
	}
}

func TestGammas(t *testing.T) {
	ζ := big.NewInt(17)
	q := big.NewInt(q)
	for k, gamma := range gammas {
		// ζ^2BitRev7(i)+1
		exp := new(big.Int).Exp(ζ, big.NewInt(int64(BitRev7(uint8(k)))*2+1), q)
		if big.NewInt(int64(gamma)).Cmp(exp) != 0 {
			t.Errorf("gammas[%d] = %v, expected %v", k, gamma, exp)
		}
	}
}

func TestRoundTrip(t *testing.T) {
	dk, err := GenerateKey()
	if err != nil {