ENH: random: Allow ngood=0 or nbad=0 in mtrand.hypergeometric.

Also edited the 'Parameters' section of the docstring to comply
with the numpy docstring standard.

3 files changed, 259 insertions, 252 deletions

diff --git a/numpy/random/mtrand/mtrand.c b/numpy/random/mtrand/mtrand.c
index a81170674..c411298de 100644
--- a/numpy/random/mtrand/mtrand.c
+++ b/numpy/random/mtrand/mtrand.c
@@ -1,4 +1,4 @@
-/* Generated by Cython 0.19.1 on Mon May 27 11:39:14 2013 */
+/* Generated by Cython 0.19.1 on Thu Jun 13 06:23:57 2013 */
 
 #define PY_SSIZE_T_CLEAN
 #ifndef CYTHON_USE_PYLONG_INTERNALS
@@ -264,7 +264,7 @@
 #ifndef CYTHON_RESTRICT
   #if defined(__GNUC__)
     #define CYTHON_RESTRICT __restrict__
-  #elif defined(_MSC_VER) && _MSC_VER >= 1400
+  #elif defined(_MSC_VER)
     #define CYTHON_RESTRICT __restrict
   #elif defined (__STDC_VERSION__) && __STDC_VERSION__ >= 199901L
     #define CYTHON_RESTRICT restrict
@@ -854,11 +854,11 @@ static int __Pyx_InitStrings(__Pyx_StringTabEntry *t); /*proto*/
 /* Module declarations from 'numpy' */
 
 /* Module declarations from 'mtrand' */
-static PyTypeObject *__pyx_ptype_6mtrand_ndarray = 0;
-static PyTypeObject *__pyx_ptype_6mtrand_broadcast = 0;
 static PyTypeObject *__pyx_ptype_6mtrand_RandomState = 0;
 static PyTypeObject *__pyx_ptype_6mtrand_flatiter = 0;
+static PyTypeObject *__pyx_ptype_6mtrand_ndarray = 0;
 static PyTypeObject *__pyx_ptype_6mtrand_dtype = 0;
+static PyTypeObject *__pyx_ptype_6mtrand_broadcast = 0;
 static PyObject *__pyx_f_6mtrand_cont0_array(rk_state *, __pyx_t_6mtrand_rk_cont0, PyObject *); /*proto*/
 static PyObject *__pyx_f_6mtrand_cont1_array_sc(rk_state *, __pyx_t_6mtrand_rk_cont1, PyObject *, double); /*proto*/
 static PyObject *__pyx_f_6mtrand_cont1_array(rk_state *, __pyx_t_6mtrand_rk_cont1, PyObject *, PyArrayObject *); /*proto*/
@@ -985,8 +985,8 @@ static char __pyx_k_156[] = "lam value too large.";
 static char __pyx_k_158[] = "a <= 1.0";
 static char __pyx_k_161[] = "p < 0.0";
 static char __pyx_k_163[] = "p > 1.0";
-static char __pyx_k_167[] = "ngood < 1";
-static char __pyx_k_169[] = "nbad < 1";
+static char __pyx_k_167[] = "ngood < 0";
+static char __pyx_k_169[] = "nbad < 0";
 static char __pyx_k_171[] = "nsample < 1";
 static char __pyx_k_173[] = "ngood + nbad < nsample";
 static char __pyx_k_179[] = "p <= 0.0";
@@ -1071,7 +1071,7 @@ static char __pyx_k_269[] = "\n        zipf(a, size=None)\n\n        Draw sample
 static char __pyx_k_270[] = "RandomState.geometric (line 3770)";
 static char __pyx_k_271[] = "\n        geometric(p, size=None)\n\n        Draw samples from the geometric distribution.\n\n        Bernoulli trials are experiments with one of two outcomes:\n        success or failure (an example of such an experiment is flipping\n        a coin).  The geometric distribution models the number of trials\n        that must be run in order to achieve success.  It is therefore\n        supported on the positive integers, ``k = 1, 2, ...``.\n\n        The probability mass function of the geometric distribution is\n\n        .. math:: f(k) = (1 - p)^{k - 1} p\n\n        where `p` is the probability of success of an individual trial.\n\n        Parameters\n        ----------\n        p : float\n            The probability of success of an individual trial.\n        size : tuple of ints\n            Number of values to draw from the distribution.  The output\n            is shaped according to `size`.\n\n        Returns\n        -------\n        out : ndarray\n            Samples from the geometric distribution, shaped according to\n            `size`.\n\n        Examples\n        --------\n        Draw ten thousand values from the geometric distribution,\n        with the probability of an individual success equal to 0.35:\n\n        >>> z = np.random.geometric(p=0.35, size=10000)\n\n        How many trials succeeded after a single run?\n\n        >>> (z == 1).sum() / 10000.\n        0.34889999999999999 #random\n\n        ";
 static char __pyx_k_272[] = "RandomState.hypergeometric (line 3836)";
-static char __pyx_k_273[] = "\n        hypergeometric(ngood, nbad, nsample, size=None)\n\n        Draw samples from a Hypergeometric distribution.\n\n        Samples are drawn from a Hypergeometric distribution with specified\n        parameters, ngood (ways to make a good selection), nbad (ways to make\n        a bad selection), and nsample = number of items sampled, which is less\n        than or equal to the sum ngood + nbad.\n\n        Parameters\n        ----------\n        ngood : float (but truncated to an integer)\n                parameter, > 0.\n        nbad  : float\n                parameter, >= 0.\n        nsample  : float\n                   parameter, > 0 and <= ngood+nbad\n        size : {tuple, int}\n            Output shape.  If the given shape is, e.g., ``(m, n, k)``, then\n            ``m * n * k`` samples are drawn.\n\n        Returns\n        -------\n        samples : {ndarray, scalar}\n                  where the values are all integers in  [0, n].\n\n        See Also\n        --------\n        scipy.stats.distributions.hypergeom : probability density function,\n            distribution or cumulative density function, etc.\n\n        Notes\n        -----\n        The probability density for the Hypergeometric distribution is\n\n        .. math:: P(x) = \\frac{\\binom{m}{n}\\binom{N-m}{n-x}}{\\binom{N}{n}},\n\n        where :math:`0 \\le x \\le m` and :math:`n+m-N \\le x \\le n`\n\n        for P(x) the probability of x successes, n = ngood, m = nbad, and\n        N = number of samples.\n\n        Consider an urn with black and white marbles in it, ngood of them\n        black and nbad are white. If you draw nsample balls without\n        replacement, then the Hypergeometric distribution describes the\n        distribution of black balls in the drawn sample.\n\n        Note that this distribution is very similar to the Binomial\n        distribution, except that in this case, samples are drawn without\n        replacement, whereas in the Binomial case samples are drawn wit""h\n        replacement (or the sample space is infinite). As the sample space\n        becomes large, this distribution approaches the Binomial.\n\n        References\n        ----------\n        .. [1] Lentner, Marvin, \"Elementary Applied Statistics\", Bogden\n               and Quigley, 1972.\n        .. [2] Weisstein, Eric W. \"Hypergeometric Distribution.\" From\n               MathWorld--A Wolfram Web Resource.\n               http://mathworld.wolfram.com/HypergeometricDistribution.html\n        .. [3] Wikipedia, \"Hypergeometric-distribution\",\n               http://en.wikipedia.org/wiki/Hypergeometric-distribution\n\n        Examples\n        --------\n        Draw samples from the distribution:\n\n        >>> ngood, nbad, nsamp = 100, 2, 10\n        # number of good, number of bad, and number of samples\n        >>> s = np.random.hypergeometric(ngood, nbad, nsamp, 1000)\n        >>> hist(s)\n        #   note that it is very unlikely to grab both bad items\n\n        Suppose you have an urn with 15 white and 15 black marbles.\n        If you pull 15 marbles at random, how likely is it that\n        12 or more of them are one color?\n\n        >>> s = np.random.hypergeometric(15, 15, 15, 100000)\n        >>> sum(s>=12)/100000. + sum(s<=3)/100000.\n        #   answer = 0.003 ... pretty unlikely!\n\n        ";
+static char __pyx_k_273[] = "\n        hypergeometric(ngood, nbad, nsample, size=None)\n\n        Draw samples from a Hypergeometric distribution.\n\n        Samples are drawn from a Hypergeometric distribution with specified\n        parameters, ngood (ways to make a good selection), nbad (ways to make\n        a bad selection), and nsample = number of items sampled, which is less\n        than or equal to the sum ngood + nbad.\n\n        Parameters\n        ----------\n        ngood : int or array_like\n            Number of ways to make a good selection.  Must be nonnegative.\n        nbad : int or array_like\n            Number of ways to make a bad selection.  Must be nonnegative.\n        nsample : int or array_like\n            Number of items sampled.  Must be at least 1 and at most\n            ``ngood + nbad``.\n        size : int or tuple of int\n            Output shape.  If the given shape is, e.g., ``(m, n, k)``, then\n            ``m * n * k`` samples are drawn.\n\n        Returns\n        -------\n        samples : ndarray or scalar\n            The values are all integers in  [0, n].\n\n        See Also\n        --------\n        scipy.stats.distributions.hypergeom : probability density function,\n            distribution or cumulative density function, etc.\n\n        Notes\n        -----\n        The probability density for the Hypergeometric distribution is\n\n        .. math:: P(x) = \\frac{\\binom{m}{n}\\binom{N-m}{n-x}}{\\binom{N}{n}},\n\n        where :math:`0 \\le x \\le m` and :math:`n+m-N \\le x \\le n`\n\n        for P(x) the probability of x successes, n = ngood, m = nbad, and\n        N = number of samples.\n\n        Consider an urn with black and white marbles in it, ngood of them\n        black and nbad are white. If you draw nsample balls without\n        replacement, then the Hypergeometric distribution describes the\n        distribution of black balls in the drawn sample.\n\n        Note that this distribution is very similar to the Binomial\n        distrib""ution, except that in this case, samples are drawn without\n        replacement, whereas in the Binomial case samples are drawn with\n        replacement (or the sample space is infinite). As the sample space\n        becomes large, this distribution approaches the Binomial.\n\n        References\n        ----------\n        .. [1] Lentner, Marvin, \"Elementary Applied Statistics\", Bogden\n               and Quigley, 1972.\n        .. [2] Weisstein, Eric W. \"Hypergeometric Distribution.\" From\n               MathWorld--A Wolfram Web Resource.\n               http://mathworld.wolfram.com/HypergeometricDistribution.html\n        .. [3] Wikipedia, \"Hypergeometric-distribution\",\n               http://en.wikipedia.org/wiki/Hypergeometric-distribution\n\n        Examples\n        --------\n        Draw samples from the distribution:\n\n        >>> ngood, nbad, nsamp = 100, 2, 10\n        # number of good, number of bad, and number of samples\n        >>> s = np.random.hypergeometric(ngood, nbad, nsamp, 1000)\n        >>> hist(s)\n        #   note that it is very unlikely to grab both bad items\n\n        Suppose you have an urn with 15 white and 15 black marbles.\n        If you pull 15 marbles at random, how likely is it that\n        12 or more of them are one color?\n\n        >>> s = np.random.hypergeometric(15, 15, 15, 100000)\n        >>> sum(s>=12)/100000. + sum(s<=3)/100000.\n        #   answer = 0.003 ... pretty unlikely!\n\n        ";
 static char __pyx_k_274[] = "RandomState.logseries (line 3955)";
 static char __pyx_k_275[] = "\n        logseries(p, size=None)\n\n        Draw samples from a Logarithmic Series distribution.\n\n        Samples are drawn from a Log Series distribution with specified\n        parameter, p (probability, 0 < p < 1).\n\n        Parameters\n        ----------\n        loc : float\n\n        scale : float > 0.\n\n        size : {tuple, int}\n            Output shape.  If the given shape is, e.g., ``(m, n, k)``, then\n            ``m * n * k`` samples are drawn.\n\n        Returns\n        -------\n        samples : {ndarray, scalar}\n                  where the values are all integers in  [0, n].\n\n        See Also\n        --------\n        scipy.stats.distributions.logser : probability density function,\n            distribution or cumulative density function, etc.\n\n        Notes\n        -----\n        The probability density for the Log Series distribution is\n\n        .. math:: P(k) = \\frac{-p^k}{k \\ln(1-p)},\n\n        where p = probability.\n\n        The Log Series distribution is frequently used to represent species\n        richness and occurrence, first proposed by Fisher, Corbet, and\n        Williams in 1943 [2].  It may also be used to model the numbers of\n        occupants seen in cars [3].\n\n        References\n        ----------\n        .. [1] Buzas, Martin A.; Culver, Stephen J.,  Understanding regional\n               species diversity through the log series distribution of\n               occurrences: BIODIVERSITY RESEARCH Diversity & Distributions,\n               Volume 5, Number 5, September 1999 , pp. 187-195(9).\n        .. [2] Fisher, R.A,, A.S. Corbet, and C.B. Williams. 1943. The\n               relation between the number of species and the number of\n               individuals in a random sample of an animal population.\n               Journal of Animal Ecology, 12:42-58.\n        .. [3] D. J. Hand, F. Daly, D. Lunn, E. Ostrowski, A Handbook of Small\n               Data Sets, CRC Press, 1994.\n        .. [4] Wikipedia, \"Log""arithmic-distribution\",\n               http://en.wikipedia.org/wiki/Logarithmic-distribution\n\n        Examples\n        --------\n        Draw samples from the distribution:\n\n        >>> a = .6\n        >>> s = np.random.logseries(a, 10000)\n        >>> count, bins, ignored = plt.hist(s)\n\n        #   plot against distribution\n\n        >>> def logseries(k, p):\n        ...     return -p**k/(k*log(1-p))\n        >>> plt.plot(bins, logseries(bins, a)*count.max()/\n                     logseries(bins, a).max(), 'r')\n        >>> plt.show()\n\n        ";
 static char __pyx_k_276[] = "RandomState.multivariate_normal (line 4050)";
@@ -18225,7 +18225,7 @@ static PyObject *__pyx_pf_6mtrand_11RandomState_90geometric(struct __pyx_obj_6mt
 
 /* Python wrapper */
 static PyObject *__pyx_pw_6mtrand_11RandomState_93hypergeometric(PyObject *__pyx_v_self, PyObject *__pyx_args, PyObject *__pyx_kwds); /*proto*/
-static char __pyx_doc_6mtrand_11RandomState_92hypergeometric[] = "\n        hypergeometric(ngood, nbad, nsample, size=None)\n\n        Draw samples from a Hypergeometric distribution.\n\n        Samples are drawn from a Hypergeometric distribution with specified\n        parameters, ngood (ways to make a good selection), nbad (ways to make\n        a bad selection), and nsample = number of items sampled, which is less\n        than or equal to the sum ngood + nbad.\n\n        Parameters\n        ----------\n        ngood : float (but truncated to an integer)\n                parameter, > 0.\n        nbad  : float\n                parameter, >= 0.\n        nsample  : float\n                   parameter, > 0 and <= ngood+nbad\n        size : {tuple, int}\n            Output shape.  If the given shape is, e.g., ``(m, n, k)``, then\n            ``m * n * k`` samples are drawn.\n\n        Returns\n        -------\n        samples : {ndarray, scalar}\n                  where the values are all integers in  [0, n].\n\n        See Also\n        --------\n        scipy.stats.distributions.hypergeom : probability density function,\n            distribution or cumulative density function, etc.\n\n        Notes\n        -----\n        The probability density for the Hypergeometric distribution is\n\n        .. math:: P(x) = \\frac{\\binom{m}{n}\\binom{N-m}{n-x}}{\\binom{N}{n}},\n\n        where :math:`0 \\le x \\le m` and :math:`n+m-N \\le x \\le n`\n\n        for P(x) the probability of x successes, n = ngood, m = nbad, and\n        N = number of samples.\n\n        Consider an urn with black and white marbles in it, ngood of them\n        black and nbad are white. If you draw nsample balls without\n        replacement, then the Hypergeometric distribution describes the\n        distribution of black balls in the drawn sample.\n\n        Note that this distribution is very similar to the Binomial\n        distribution, except that in this case, samples are drawn without\n        replacement, whereas in the Binomial case samples are drawn wit""h\n        replacement (or the sample space is infinite). As the sample space\n        becomes large, this distribution approaches the Binomial.\n\n        References\n        ----------\n        .. [1] Lentner, Marvin, \"Elementary Applied Statistics\", Bogden\n               and Quigley, 1972.\n        .. [2] Weisstein, Eric W. \"Hypergeometric Distribution.\" From\n               MathWorld--A Wolfram Web Resource.\n               http://mathworld.wolfram.com/HypergeometricDistribution.html\n        .. [3] Wikipedia, \"Hypergeometric-distribution\",\n               http://en.wikipedia.org/wiki/Hypergeometric-distribution\n\n        Examples\n        --------\n        Draw samples from the distribution:\n\n        >>> ngood, nbad, nsamp = 100, 2, 10\n        # number of good, number of bad, and number of samples\n        >>> s = np.random.hypergeometric(ngood, nbad, nsamp, 1000)\n        >>> hist(s)\n        #   note that it is very unlikely to grab both bad items\n\n        Suppose you have an urn with 15 white and 15 black marbles.\n        If you pull 15 marbles at random, how likely is it that\n        12 or more of them are one color?\n\n        >>> s = np.random.hypergeometric(15, 15, 15, 100000)\n        >>> sum(s>=12)/100000. + sum(s<=3)/100000.\n        #   answer = 0.003 ... pretty unlikely!\n\n        ";
+static char __pyx_doc_6mtrand_11RandomState_92hypergeometric[] = "\n        hypergeometric(ngood, nbad, nsample, size=None)\n\n        Draw samples from a Hypergeometric distribution.\n\n        Samples are drawn from a Hypergeometric distribution with specified\n        parameters, ngood (ways to make a good selection), nbad (ways to make\n        a bad selection), and nsample = number of items sampled, which is less\n        than or equal to the sum ngood + nbad.\n\n        Parameters\n        ----------\n        ngood : int or array_like\n            Number of ways to make a good selection.  Must be nonnegative.\n        nbad : int or array_like\n            Number of ways to make a bad selection.  Must be nonnegative.\n        nsample : int or array_like\n            Number of items sampled.  Must be at least 1 and at most\n            ``ngood + nbad``.\n        size : int or tuple of int\n            Output shape.  If the given shape is, e.g., ``(m, n, k)``, then\n            ``m * n * k`` samples are drawn.\n\n        Returns\n        -------\n        samples : ndarray or scalar\n            The values are all integers in  [0, n].\n\n        See Also\n        --------\n        scipy.stats.distributions.hypergeom : probability density function,\n            distribution or cumulative density function, etc.\n\n        Notes\n        -----\n        The probability density for the Hypergeometric distribution is\n\n        .. math:: P(x) = \\frac{\\binom{m}{n}\\binom{N-m}{n-x}}{\\binom{N}{n}},\n\n        where :math:`0 \\le x \\le m` and :math:`n+m-N \\le x \\le n`\n\n        for P(x) the probability of x successes, n = ngood, m = nbad, and\n        N = number of samples.\n\n        Consider an urn with black and white marbles in it, ngood of them\n        black and nbad are white. If you draw nsample balls without\n        replacement, then the Hypergeometric distribution describes the\n        distribution of black balls in the drawn sample.\n\n        Note that this distribution is very similar to the Binomial\n        distrib""ution, except that in this case, samples are drawn without\n        replacement, whereas in the Binomial case samples are drawn with\n        replacement (or the sample space is infinite). As the sample space\n        becomes large, this distribution approaches the Binomial.\n\n        References\n        ----------\n        .. [1] Lentner, Marvin, \"Elementary Applied Statistics\", Bogden\n               and Quigley, 1972.\n        .. [2] Weisstein, Eric W. \"Hypergeometric Distribution.\" From\n               MathWorld--A Wolfram Web Resource.\n               http://mathworld.wolfram.com/HypergeometricDistribution.html\n        .. [3] Wikipedia, \"Hypergeometric-distribution\",\n               http://en.wikipedia.org/wiki/Hypergeometric-distribution\n\n        Examples\n        --------\n        Draw samples from the distribution:\n\n        >>> ngood, nbad, nsamp = 100, 2, 10\n        # number of good, number of bad, and number of samples\n        >>> s = np.random.hypergeometric(ngood, nbad, nsamp, 1000)\n        >>> hist(s)\n        #   note that it is very unlikely to grab both bad items\n\n        Suppose you have an urn with 15 white and 15 black marbles.\n        If you pull 15 marbles at random, how likely is it that\n        12 or more of them are one color?\n\n        >>> s = np.random.hypergeometric(15, 15, 15, 100000)\n        >>> sum(s>=12)/100000. + sum(s<=3)/100000.\n        #   answer = 0.003 ... pretty unlikely!\n\n        ";
 static PyObject *__pyx_pw_6mtrand_11RandomState_93hypergeometric(PyObject *__pyx_v_self, PyObject *__pyx_args, PyObject *__pyx_kwds) {
   PyObject *__pyx_v_ngood = 0;
   PyObject *__pyx_v_nbad = 0;
@@ -18332,7 +18332,7 @@ static PyObject *__pyx_pf_6mtrand_11RandomState_92hypergeometric(struct __pyx_ob
   int __pyx_clineno = 0;
   __Pyx_RefNannySetupContext("hypergeometric", 0);
 
-  /* "mtrand.pyx":3923
+  /* "mtrand.pyx":3924
  *         cdef long lngood, lnbad, lnsample
  * 
  *         lngood = PyInt_AsLong(ngood)             # <<<<<<<<<<<<<<
@@ -18341,7 +18341,7 @@ static PyObject *__pyx_pf_6mtrand_11RandomState_92hypergeometric(struct __pyx_ob
  */
   __pyx_v_lngood = PyInt_AsLong(__pyx_v_ngood);
 
-  /* "mtrand.pyx":3924
+  /* "mtrand.pyx":3925
  * 
  *         lngood = PyInt_AsLong(ngood)
  *         lnbad = PyInt_AsLong(nbad)             # <<<<<<<<<<<<<<
@@ -18350,142 +18350,131 @@ static PyObject *__pyx_pf_6mtrand_11RandomState_92hypergeometric(struct __pyx_ob
  */
   __pyx_v_lnbad = PyInt_AsLong(__pyx_v_nbad);
 
-  /* "mtrand.pyx":3925
+  /* "mtrand.pyx":3926
  *         lngood = PyInt_AsLong(ngood)
  *         lnbad = PyInt_AsLong(nbad)
  *         lnsample = PyInt_AsLong(nsample)             # <<<<<<<<<<<<<<
  *         if not PyErr_Occurred():
- *             if ngood < 1:
+ *             if lngood < 0:
  */
   __pyx_v_lnsample = PyInt_AsLong(__pyx_v_nsample);
 
-  /* "mtrand.pyx":3926
+  /* "mtrand.pyx":3927
  *         lnbad = PyInt_AsLong(nbad)
  *         lnsample = PyInt_AsLong(nsample)
  *         if not PyErr_Occurred():             # <<<<<<<<<<<<<<
- *             if ngood < 1:
- *                 raise ValueError("ngood < 1")
+ *             if lngood < 0:
+ *                 raise ValueError("ngood < 0")
  */
   __pyx_t_1 = ((!(PyErr_Occurred() != 0)) != 0);
   if (__pyx_t_1) {
 
-    /* "mtrand.pyx":3927
+    /* "mtrand.pyx":3928
  *         lnsample = PyInt_AsLong(nsample)
  *         if not PyErr_Occurred():
- *             if ngood < 1:             # <<<<<<<<<<<<<<
- *                 raise ValueError("ngood < 1")
- *             if nbad < 1:
+ *             if lngood < 0:             # <<<<<<<<<<<<<<
+ *                 raise ValueError("ngood < 0")
+ *             if lnbad < 0:
  */
-    __pyx_t_2 = PyObject_RichCompare(__pyx_v_ngood, __pyx_int_1, Py_LT); __Pyx_XGOTREF(__pyx_t_2); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3927; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
-    __pyx_t_1 = __Pyx_PyObject_IsTrue(__pyx_t_2); if (unlikely(__pyx_t_1 < 0)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3927; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
-    __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;
+    __pyx_t_1 = ((__pyx_v_lngood < 0) != 0);
     if (__pyx_t_1) {
 
-      /* "mtrand.pyx":3928
+      /* "mtrand.pyx":3929
  *         if not PyErr_Occurred():
- *             if ngood < 1:
- *                 raise ValueError("ngood < 1")             # <<<<<<<<<<<<<<
- *             if nbad < 1:
- *                 raise ValueError("nbad < 1")
+ *             if lngood < 0:
+ *                 raise ValueError("ngood < 0")             # <<<<<<<<<<<<<<
+ *             if lnbad < 0:
+ *                 raise ValueError("nbad < 0")
  */
-      __pyx_t_2 = PyObject_Call(__pyx_builtin_ValueError, ((PyObject *)__pyx_k_tuple_168), NULL); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3928; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+      __pyx_t_2 = PyObject_Call(__pyx_builtin_ValueError, ((PyObject *)__pyx_k_tuple_168), NULL); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3929; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
       __Pyx_GOTREF(__pyx_t_2);
       __Pyx_Raise(__pyx_t_2, 0, 0, 0);
       __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;
-      {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3928; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+      {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3929; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
       goto __pyx_L4;
     }
     __pyx_L4:;
 
-    /* "mtrand.pyx":3929
- *             if ngood < 1:
- *                 raise ValueError("ngood < 1")
- *             if nbad < 1:             # <<<<<<<<<<<<<<
- *                 raise ValueError("nbad < 1")
- *             if nsample < 1:
+    /* "mtrand.pyx":3930
+ *             if lngood < 0:
+ *                 raise ValueError("ngood < 0")
+ *             if lnbad < 0:             # <<<<<<<<<<<<<<
+ *                 raise ValueError("nbad < 0")
+ *             if lnsample < 1:
  */
-    __pyx_t_2 = PyObject_RichCompare(__pyx_v_nbad, __pyx_int_1, Py_LT); __Pyx_XGOTREF(__pyx_t_2); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3929; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
-    __pyx_t_1 = __Pyx_PyObject_IsTrue(__pyx_t_2); if (unlikely(__pyx_t_1 < 0)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3929; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
-    __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;
+    __pyx_t_1 = ((__pyx_v_lnbad < 0) != 0);
     if (__pyx_t_1) {
 
-      /* "mtrand.pyx":3930
- *                 raise ValueError("ngood < 1")
- *             if nbad < 1:
- *                 raise ValueError("nbad < 1")             # <<<<<<<<<<<<<<
- *             if nsample < 1:
+      /* "mtrand.pyx":3931
+ *                 raise ValueError("ngood < 0")
+ *             if lnbad < 0:
+ *                 raise ValueError("nbad < 0")             # <<<<<<<<<<<<<<
+ *             if lnsample < 1:
  *                 raise ValueError("nsample < 1")
  */
-      __pyx_t_2 = PyObject_Call(__pyx_builtin_ValueError, ((PyObject *)__pyx_k_tuple_170), NULL); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3930; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+      __pyx_t_2 = PyObject_Call(__pyx_builtin_ValueError, ((PyObject *)__pyx_k_tuple_170), NULL); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3931; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
       __Pyx_GOTREF(__pyx_t_2);
       __Pyx_Raise(__pyx_t_2, 0, 0, 0);
       __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;
-      {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3930; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+      {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3931; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
       goto __pyx_L5;
     }
     __pyx_L5:;
 
-    /* "mtrand.pyx":3931
- *             if nbad < 1:
- *                 raise ValueError("nbad < 1")
- *             if nsample < 1:             # <<<<<<<<<<<<<<
+    /* "mtrand.pyx":3932
+ *             if lnbad < 0:
+ *                 raise ValueError("nbad < 0")
+ *             if lnsample < 1:             # <<<<<<<<<<<<<<
  *                 raise ValueError("nsample < 1")
- *             if ngood + nbad < nsample:
+ *             if lngood + lnbad < lnsample:
  */
-    __pyx_t_2 = PyObject_RichCompare(__pyx_v_nsample, __pyx_int_1, Py_LT); __Pyx_XGOTREF(__pyx_t_2); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3931; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
-    __pyx_t_1 = __Pyx_PyObject_IsTrue(__pyx_t_2); if (unlikely(__pyx_t_1 < 0)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3931; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
-    __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;
+    __pyx_t_1 = ((__pyx_v_lnsample < 1) != 0);
     if (__pyx_t_1) {
 
-      /* "mtrand.pyx":3932
- *                 raise ValueError("nbad < 1")
- *             if nsample < 1:
+      /* "mtrand.pyx":3933
+ *                 raise ValueError("nbad < 0")
+ *             if lnsample < 1:
  *                 raise ValueError("nsample < 1")             # <<<<<<<<<<<<<<
- *             if ngood + nbad < nsample:
+ *             if lngood + lnbad < lnsample:
  *                 raise ValueError("ngood + nbad < nsample")
  */
-      __pyx_t_2 = PyObject_Call(__pyx_builtin_ValueError, ((PyObject *)__pyx_k_tuple_172), NULL); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3932; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+      __pyx_t_2 = PyObject_Call(__pyx_builtin_ValueError, ((PyObject *)__pyx_k_tuple_172), NULL); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3933; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
       __Pyx_GOTREF(__pyx_t_2);
       __Pyx_Raise(__pyx_t_2, 0, 0, 0);
       __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;
-      {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3932; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+      {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3933; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
       goto __pyx_L6;
     }
     __pyx_L6:;
 
-    /* "mtrand.pyx":3933
- *             if nsample < 1:
+    /* "mtrand.pyx":3934
+ *             if lnsample < 1:
  *                 raise ValueError("nsample < 1")
- *             if ngood + nbad < nsample:             # <<<<<<<<<<<<<<
+ *             if lngood + lnbad < lnsample:             # <<<<<<<<<<<<<<
  *                 raise ValueError("ngood + nbad < nsample")
  *             return discnmN_array_sc(self.internal_state, rk_hypergeometric, size,
  */
-    __pyx_t_2 = PyNumber_Add(__pyx_v_ngood, __pyx_v_nbad); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3933; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
-    __Pyx_GOTREF(__pyx_t_2);
-    __pyx_t_3 = PyObject_RichCompare(__pyx_t_2, __pyx_v_nsample, Py_LT); __Pyx_XGOTREF(__pyx_t_3); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3933; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
-    __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;
-    __pyx_t_1 = __Pyx_PyObject_IsTrue(__pyx_t_3); if (unlikely(__pyx_t_1 < 0)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3933; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
-    __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
+    __pyx_t_1 = (((__pyx_v_lngood + __pyx_v_lnbad) < __pyx_v_lnsample) != 0);
     if (__pyx_t_1) {
 
-      /* "mtrand.pyx":3934
+      /* "mtrand.pyx":3935
  *                 raise ValueError("nsample < 1")
- *             if ngood + nbad < nsample:
+ *             if lngood + lnbad < lnsample:
  *                 raise ValueError("ngood + nbad < nsample")             # <<<<<<<<<<<<<<
  *             return discnmN_array_sc(self.internal_state, rk_hypergeometric, size,
  *                                     lngood, lnbad, lnsample)
  */
-      __pyx_t_3 = PyObject_Call(__pyx_builtin_ValueError, ((PyObject *)__pyx_k_tuple_174), NULL); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3934; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
-      __Pyx_GOTREF(__pyx_t_3);
-      __Pyx_Raise(__pyx_t_3, 0, 0, 0);
-      __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
-      {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3934; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+      __pyx_t_2 = PyObject_Call(__pyx_builtin_ValueError, ((PyObject *)__pyx_k_tuple_174), NULL); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3935; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+      __Pyx_GOTREF(__pyx_t_2);
+      __Pyx_Raise(__pyx_t_2, 0, 0, 0);
+      __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;
+      {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3935; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
       goto __pyx_L7;
     }
     __pyx_L7:;
 
-    /* "mtrand.pyx":3935
- *             if ngood + nbad < nsample:
+    /* "mtrand.pyx":3936
+ *             if lngood + lnbad < lnsample:
  *                 raise ValueError("ngood + nbad < nsample")
  *             return discnmN_array_sc(self.internal_state, rk_hypergeometric, size,             # <<<<<<<<<<<<<<
  *                                     lngood, lnbad, lnsample)
@@ -18493,24 +18482,24 @@ static PyObject *__pyx_pf_6mtrand_11RandomState_92hypergeometric(struct __pyx_ob
  */
     __Pyx_XDECREF(__pyx_r);
 
-    /* "mtrand.pyx":3936
+    /* "mtrand.pyx":3937
  *                 raise ValueError("ngood + nbad < nsample")
  *             return discnmN_array_sc(self.internal_state, rk_hypergeometric, size,
  *                                     lngood, lnbad, lnsample)             # <<<<<<<<<<<<<<
  * 
- * 
+ *         PyErr_Clear()
  */
-    __pyx_t_3 = __pyx_f_6mtrand_discnmN_array_sc(__pyx_v_self->internal_state, rk_hypergeometric, __pyx_v_size, __pyx_v_lngood, __pyx_v_lnbad, __pyx_v_lnsample); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3935; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
-    __Pyx_GOTREF(__pyx_t_3);
-    __pyx_r = __pyx_t_3;
-    __pyx_t_3 = 0;
+    __pyx_t_2 = __pyx_f_6mtrand_discnmN_array_sc(__pyx_v_self->internal_state, rk_hypergeometric, __pyx_v_size, __pyx_v_lngood, __pyx_v_lnbad, __pyx_v_lnsample); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3936; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+    __Pyx_GOTREF(__pyx_t_2);
+    __pyx_r = __pyx_t_2;
+    __pyx_t_2 = 0;
     goto __pyx_L0;
     goto __pyx_L3;
   }
   __pyx_L3:;
 
   /* "mtrand.pyx":3939
- * 
+ *                                     lngood, lnbad, lnsample)
  * 
  *         PyErr_Clear()             # <<<<<<<<<<<<<<
  * 
@@ -18525,92 +18514,92 @@ static PyObject *__pyx_pf_6mtrand_11RandomState_92hypergeometric(struct __pyx_ob
  *         onbad = <ndarray>PyArray_FROM_OTF(nbad, NPY_LONG, NPY_ARRAY_ALIGNED)
  *         onsample = <ndarray>PyArray_FROM_OTF(nsample, NPY_LONG, NPY_ARRAY_ALIGNED)
  */
-  __pyx_t_3 = PyArray_FROM_OTF(__pyx_v_ngood, NPY_LONG, NPY_ARRAY_ALIGNED); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3941; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
-  __Pyx_GOTREF(__pyx_t_3);
-  __pyx_t_2 = __pyx_t_3;
-  __Pyx_INCREF(__pyx_t_2);
-  __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
-  __pyx_v_ongood = ((PyArrayObject *)__pyx_t_2);
-  __pyx_t_2 = 0;
-
-  /* "mtrand.pyx":3942
- * 
- *         ongood = <ndarray>PyArray_FROM_OTF(ngood, NPY_LONG, NPY_ARRAY_ALIGNED)
- *         onbad = <ndarray>PyArray_FROM_OTF(nbad, NPY_LONG, NPY_ARRAY_ALIGNED)             # <<<<<<<<<<<<<<
- *         onsample = <ndarray>PyArray_FROM_OTF(nsample, NPY_LONG, NPY_ARRAY_ALIGNED)
- *         if np.any(np.less(ongood, 1)):
- */
-  __pyx_t_2 = PyArray_FROM_OTF(__pyx_v_nbad, NPY_LONG, NPY_ARRAY_ALIGNED); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3942; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+  __pyx_t_2 = PyArray_FROM_OTF(__pyx_v_ngood, NPY_LONG, NPY_ARRAY_ALIGNED); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3941; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
   __Pyx_GOTREF(__pyx_t_2);
   __pyx_t_3 = __pyx_t_2;
   __Pyx_INCREF(__pyx_t_3);
   __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;
-  __pyx_v_onbad = ((PyArrayObject *)__pyx_t_3);
+  __pyx_v_ongood = ((PyArrayObject *)__pyx_t_3);
   __pyx_t_3 = 0;
 
-  /* "mtrand.pyx":3943
+  /* "mtrand.pyx":3942
+ * 
  *         ongood = <ndarray>PyArray_FROM_OTF(ngood, NPY_LONG, NPY_ARRAY_ALIGNED)
- *         onbad = <ndarray>PyArray_FROM_OTF(nbad, NPY_LONG, NPY_ARRAY_ALIGNED)
- *         onsample = <ndarray>PyArray_FROM_OTF(nsample, NPY_LONG, NPY_ARRAY_ALIGNED)             # <<<<<<<<<<<<<<
- *         if np.any(np.less(ongood, 1)):
- *             raise ValueError("ngood < 1")
+ *         onbad = <ndarray>PyArray_FROM_OTF(nbad, NPY_LONG, NPY_ARRAY_ALIGNED)             # <<<<<<<<<<<<<<
+ *         onsample = <ndarray>PyArray_FROM_OTF(nsample, NPY_LONG, NPY_ARRAY_ALIGNED)
+ *         if np.any(np.less(ongood, 0)):
  */
-  __pyx_t_3 = PyArray_FROM_OTF(__pyx_v_nsample, NPY_LONG, NPY_ARRAY_ALIGNED); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3943; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+  __pyx_t_3 = PyArray_FROM_OTF(__pyx_v_nbad, NPY_LONG, NPY_ARRAY_ALIGNED); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3942; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
   __Pyx_GOTREF(__pyx_t_3);
   __pyx_t_2 = __pyx_t_3;
   __Pyx_INCREF(__pyx_t_2);
   __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
-  __pyx_v_onsample = ((PyArrayObject *)__pyx_t_2);
+  __pyx_v_onbad = ((PyArrayObject *)__pyx_t_2);
   __pyx_t_2 = 0;
 
+  /* "mtrand.pyx":3943
+ *         ongood = <ndarray>PyArray_FROM_OTF(ngood, NPY_LONG, NPY_ARRAY_ALIGNED)
+ *         onbad = <ndarray>PyArray_FROM_OTF(nbad, NPY_LONG, NPY_ARRAY_ALIGNED)
+ *         onsample = <ndarray>PyArray_FROM_OTF(nsample, NPY_LONG, NPY_ARRAY_ALIGNED)             # <<<<<<<<<<<<<<
+ *         if np.any(np.less(ongood, 0)):
+ *             raise ValueError("ngood < 0")
+ */
+  __pyx_t_2 = PyArray_FROM_OTF(__pyx_v_nsample, NPY_LONG, NPY_ARRAY_ALIGNED); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3943; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+  __Pyx_GOTREF(__pyx_t_2);
+  __pyx_t_3 = __pyx_t_2;
+  __Pyx_INCREF(__pyx_t_3);
+  __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;
+  __pyx_v_onsample = ((PyArrayObject *)__pyx_t_3);
+  __pyx_t_3 = 0;
+
   /* "mtrand.pyx":3944
  *         onbad = <ndarray>PyArray_FROM_OTF(nbad, NPY_LONG, NPY_ARRAY_ALIGNED)
  *         onsample = <ndarray>PyArray_FROM_OTF(nsample, NPY_LONG, NPY_ARRAY_ALIGNED)
- *         if np.any(np.less(ongood, 1)):             # <<<<<<<<<<<<<<
- *             raise ValueError("ngood < 1")
- *         if np.any(np.less(onbad, 1)):
+ *         if np.any(np.less(ongood, 0)):             # <<<<<<<<<<<<<<
+ *             raise ValueError("ngood < 0")
+ *         if np.any(np.less(onbad, 0)):
  */
-  __pyx_t_2 = __Pyx_GetModuleGlobalName(__pyx_n_s__np); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3944; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
-  __Pyx_GOTREF(__pyx_t_2);
-  __pyx_t_3 = __Pyx_PyObject_GetAttrStr(__pyx_t_2, __pyx_n_s__any); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3944; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+  __pyx_t_3 = __Pyx_GetModuleGlobalName(__pyx_n_s__np); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3944; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
   __Pyx_GOTREF(__pyx_t_3);
-  __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;
-  __pyx_t_2 = __Pyx_GetModuleGlobalName(__pyx_n_s__np); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3944; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+  __pyx_t_2 = __Pyx_PyObject_GetAttrStr(__pyx_t_3, __pyx_n_s__any); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3944; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
   __Pyx_GOTREF(__pyx_t_2);
-  __pyx_t_4 = __Pyx_PyObject_GetAttrStr(__pyx_t_2, __pyx_n_s__less); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3944; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+  __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
+  __pyx_t_3 = __Pyx_GetModuleGlobalName(__pyx_n_s__np); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3944; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+  __Pyx_GOTREF(__pyx_t_3);
+  __pyx_t_4 = __Pyx_PyObject_GetAttrStr(__pyx_t_3, __pyx_n_s__less); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3944; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
   __Pyx_GOTREF(__pyx_t_4);
-  __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;
-  __pyx_t_2 = PyTuple_New(2); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3944; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
-  __Pyx_GOTREF(__pyx_t_2);
+  __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
+  __pyx_t_3 = PyTuple_New(2); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3944; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+  __Pyx_GOTREF(__pyx_t_3);
   __Pyx_INCREF(((PyObject *)__pyx_v_ongood));
-  PyTuple_SET_ITEM(__pyx_t_2, 0, ((PyObject *)__pyx_v_ongood));
+  PyTuple_SET_ITEM(__pyx_t_3, 0, ((PyObject *)__pyx_v_ongood));
   __Pyx_GIVEREF(((PyObject *)__pyx_v_ongood));
-  __Pyx_INCREF(__pyx_int_1);
-  PyTuple_SET_ITEM(__pyx_t_2, 1, __pyx_int_1);
-  __Pyx_GIVEREF(__pyx_int_1);
-  __pyx_t_5 = PyObject_Call(__pyx_t_4, ((PyObject *)__pyx_t_2), NULL); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3944; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+  __Pyx_INCREF(__pyx_int_0);
+  PyTuple_SET_ITEM(__pyx_t_3, 1, __pyx_int_0);
+  __Pyx_GIVEREF(__pyx_int_0);
+  __pyx_t_5 = PyObject_Call(__pyx_t_4, ((PyObject *)__pyx_t_3), NULL); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3944; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
   __Pyx_GOTREF(__pyx_t_5);
   __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0;
-  __Pyx_DECREF(((PyObject *)__pyx_t_2)); __pyx_t_2 = 0;
-  __pyx_t_2 = PyTuple_New(1); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3944; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
-  __Pyx_GOTREF(__pyx_t_2);
-  PyTuple_SET_ITEM(__pyx_t_2, 0, __pyx_t_5);
+  __Pyx_DECREF(((PyObject *)__pyx_t_3)); __pyx_t_3 = 0;
+  __pyx_t_3 = PyTuple_New(1); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3944; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+  __Pyx_GOTREF(__pyx_t_3);
+  PyTuple_SET_ITEM(__pyx_t_3, 0, __pyx_t_5);
   __Pyx_GIVEREF(__pyx_t_5);
   __pyx_t_5 = 0;
-  __pyx_t_5 = PyObject_Call(__pyx_t_3, ((PyObject *)__pyx_t_2), NULL); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3944; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+  __pyx_t_5 = PyObject_Call(__pyx_t_2, ((PyObject *)__pyx_t_3), NULL); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3944; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
   __Pyx_GOTREF(__pyx_t_5);
-  __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
-  __Pyx_DECREF(((PyObject *)__pyx_t_2)); __pyx_t_2 = 0;
+  __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;
+  __Pyx_DECREF(((PyObject *)__pyx_t_3)); __pyx_t_3 = 0;
   __pyx_t_1 = __Pyx_PyObject_IsTrue(__pyx_t_5); if (unlikely(__pyx_t_1 < 0)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3944; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
   __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
   if (__pyx_t_1) {
 
     /* "mtrand.pyx":3945
  *         onsample = <ndarray>PyArray_FROM_OTF(nsample, NPY_LONG, NPY_ARRAY_ALIGNED)
- *         if np.any(np.less(ongood, 1)):
- *             raise ValueError("ngood < 1")             # <<<<<<<<<<<<<<
- *         if np.any(np.less(onbad, 1)):
- *             raise ValueError("nbad < 1")
+ *         if np.any(np.less(ongood, 0)):
+ *             raise ValueError("ngood < 0")             # <<<<<<<<<<<<<<
+ *         if np.any(np.less(onbad, 0)):
+ *             raise ValueError("nbad < 0")
  */
     __pyx_t_5 = PyObject_Call(__pyx_builtin_ValueError, ((PyObject *)__pyx_k_tuple_175), NULL); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3945; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
     __Pyx_GOTREF(__pyx_t_5);
@@ -18622,51 +18611,51 @@ static PyObject *__pyx_pf_6mtrand_11RandomState_92hypergeometric(struct __pyx_ob
   __pyx_L8:;
 
   /* "mtrand.pyx":3946
- *         if np.any(np.less(ongood, 1)):
- *             raise ValueError("ngood < 1")
- *         if np.any(np.less(onbad, 1)):             # <<<<<<<<<<<<<<
- *             raise ValueError("nbad < 1")
+ *         if np.any(np.less(ongood, 0)):
+ *             raise ValueError("ngood < 0")
+ *         if np.any(np.less(onbad, 0)):             # <<<<<<<<<<<<<<
+ *             raise ValueError("nbad < 0")
  *         if np.any(np.less(onsample, 1)):
  */
   __pyx_t_5 = __Pyx_GetModuleGlobalName(__pyx_n_s__np); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3946; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
   __Pyx_GOTREF(__pyx_t_5);
-  __pyx_t_2 = __Pyx_PyObject_GetAttrStr(__pyx_t_5, __pyx_n_s__any); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3946; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
-  __Pyx_GOTREF(__pyx_t_2);
+  __pyx_t_3 = __Pyx_PyObject_GetAttrStr(__pyx_t_5, __pyx_n_s__any); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3946; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+  __Pyx_GOTREF(__pyx_t_3);
   __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
   __pyx_t_5 = __Pyx_GetModuleGlobalName(__pyx_n_s__np); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3946; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
   __Pyx_GOTREF(__pyx_t_5);
-  __pyx_t_3 = __Pyx_PyObject_GetAttrStr(__pyx_t_5, __pyx_n_s__less); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3946; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
-  __Pyx_GOTREF(__pyx_t_3);
+  __pyx_t_2 = __Pyx_PyObject_GetAttrStr(__pyx_t_5, __pyx_n_s__less); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3946; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+  __Pyx_GOTREF(__pyx_t_2);
   __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
   __pyx_t_5 = PyTuple_New(2); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3946; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
   __Pyx_GOTREF(__pyx_t_5);
   __Pyx_INCREF(((PyObject *)__pyx_v_onbad));
   PyTuple_SET_ITEM(__pyx_t_5, 0, ((PyObject *)__pyx_v_onbad));
   __Pyx_GIVEREF(((PyObject *)__pyx_v_onbad));
-  __Pyx_INCREF(__pyx_int_1);
-  PyTuple_SET_ITEM(__pyx_t_5, 1, __pyx_int_1);
-  __Pyx_GIVEREF(__pyx_int_1);
-  __pyx_t_4 = PyObject_Call(__pyx_t_3, ((PyObject *)__pyx_t_5), NULL); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3946; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+  __Pyx_INCREF(__pyx_int_0);
+  PyTuple_SET_ITEM(__pyx_t_5, 1, __pyx_int_0);
+  __Pyx_GIVEREF(__pyx_int_0);
+  __pyx_t_4 = PyObject_Call(__pyx_t_2, ((PyObject *)__pyx_t_5), NULL); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3946; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
   __Pyx_GOTREF(__pyx_t_4);
-  __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
+  __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;
   __Pyx_DECREF(((PyObject *)__pyx_t_5)); __pyx_t_5 = 0;
   __pyx_t_5 = PyTuple_New(1); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3946; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
   __Pyx_GOTREF(__pyx_t_5);
   PyTuple_SET_ITEM(__pyx_t_5, 0, __pyx_t_4);
   __Pyx_GIVEREF(__pyx_t_4);
   __pyx_t_4 = 0;
-  __pyx_t_4 = PyObject_Call(__pyx_t_2, ((PyObject *)__pyx_t_5), NULL); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3946; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+  __pyx_t_4 = PyObject_Call(__pyx_t_3, ((PyObject *)__pyx_t_5), NULL); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3946; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
   __Pyx_GOTREF(__pyx_t_4);
-  __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;
+  __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
   __Pyx_DECREF(((PyObject *)__pyx_t_5)); __pyx_t_5 = 0;
   __pyx_t_1 = __Pyx_PyObject_IsTrue(__pyx_t_4); if (unlikely(__pyx_t_1 < 0)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3946; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
   __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0;
   if (__pyx_t_1) {
 
     /* "mtrand.pyx":3947
- *             raise ValueError("ngood < 1")
- *         if np.any(np.less(onbad, 1)):
- *             raise ValueError("nbad < 1")             # <<<<<<<<<<<<<<
+ *             raise ValueError("ngood < 0")
+ *         if np.any(np.less(onbad, 0)):
+ *             raise ValueError("nbad < 0")             # <<<<<<<<<<<<<<
  *         if np.any(np.less(onsample, 1)):
  *             raise ValueError("nsample < 1")
  */
@@ -18680,8 +18669,8 @@ static PyObject *__pyx_pf_6mtrand_11RandomState_92hypergeometric(struct __pyx_ob
   __pyx_L9:;
 
   /* "mtrand.pyx":3948
- *         if np.any(np.less(onbad, 1)):
- *             raise ValueError("nbad < 1")
+ *         if np.any(np.less(onbad, 0)):
+ *             raise ValueError("nbad < 0")
  *         if np.any(np.less(onsample, 1)):             # <<<<<<<<<<<<<<
  *             raise ValueError("nsample < 1")
  *         if np.any(np.less(np.add(ongood, onbad),onsample)):
@@ -18693,8 +18682,8 @@ static PyObject *__pyx_pf_6mtrand_11RandomState_92hypergeometric(struct __pyx_ob
   __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0;
   __pyx_t_4 = __Pyx_GetModuleGlobalName(__pyx_n_s__np); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3948; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
   __Pyx_GOTREF(__pyx_t_4);
-  __pyx_t_2 = __Pyx_PyObject_GetAttrStr(__pyx_t_4, __pyx_n_s__less); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3948; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
-  __Pyx_GOTREF(__pyx_t_2);
+  __pyx_t_3 = __Pyx_PyObject_GetAttrStr(__pyx_t_4, __pyx_n_s__less); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3948; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+  __Pyx_GOTREF(__pyx_t_3);
   __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0;
   __pyx_t_4 = PyTuple_New(2); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3948; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
   __Pyx_GOTREF(__pyx_t_4);
@@ -18704,34 +18693,34 @@ static PyObject *__pyx_pf_6mtrand_11RandomState_92hypergeometric(struct __pyx_ob
   __Pyx_INCREF(__pyx_int_1);
   PyTuple_SET_ITEM(__pyx_t_4, 1, __pyx_int_1);
   __Pyx_GIVEREF(__pyx_int_1);
-  __pyx_t_3 = PyObject_Call(__pyx_t_2, ((PyObject *)__pyx_t_4), NULL); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3948; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
-  __Pyx_GOTREF(__pyx_t_3);
-  __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;
+  __pyx_t_2 = PyObject_Call(__pyx_t_3, ((PyObject *)__pyx_t_4), NULL); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3948; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+  __Pyx_GOTREF(__pyx_t_2);
+  __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
   __Pyx_DECREF(((PyObject *)__pyx_t_4)); __pyx_t_4 = 0;
   __pyx_t_4 = PyTuple_New(1); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3948; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
   __Pyx_GOTREF(__pyx_t_4);
-  PyTuple_SET_ITEM(__pyx_t_4, 0, __pyx_t_3);
-  __Pyx_GIVEREF(__pyx_t_3);
-  __pyx_t_3 = 0;
-  __pyx_t_3 = PyObject_Call(__pyx_t_5, ((PyObject *)__pyx_t_4), NULL); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3948; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
-  __Pyx_GOTREF(__pyx_t_3);
+  PyTuple_SET_ITEM(__pyx_t_4, 0, __pyx_t_2);
+  __Pyx_GIVEREF(__pyx_t_2);
+  __pyx_t_2 = 0;
+  __pyx_t_2 = PyObject_Call(__pyx_t_5, ((PyObject *)__pyx_t_4), NULL); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3948; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+  __Pyx_GOTREF(__pyx_t_2);
   __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
   __Pyx_DECREF(((PyObject *)__pyx_t_4)); __pyx_t_4 = 0;
-  __pyx_t_1 = __Pyx_PyObject_IsTrue(__pyx_t_3); if (unlikely(__pyx_t_1 < 0)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3948; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
-  __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
+  __pyx_t_1 = __Pyx_PyObject_IsTrue(__pyx_t_2); if (unlikely(__pyx_t_1 < 0)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3948; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+  __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;
   if (__pyx_t_1) {
 
     /* "mtrand.pyx":3949
- *             raise ValueError("nbad < 1")
+ *             raise ValueError("nbad < 0")
  *         if np.any(np.less(onsample, 1)):
  *             raise ValueError("nsample < 1")             # <<<<<<<<<<<<<<
  *         if np.any(np.less(np.add(ongood, onbad),onsample)):
  *             raise ValueError("ngood + nbad < nsample")
  */
-    __pyx_t_3 = PyObject_Call(__pyx_builtin_ValueError, ((PyObject *)__pyx_k_tuple_177), NULL); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3949; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
-    __Pyx_GOTREF(__pyx_t_3);
-    __Pyx_Raise(__pyx_t_3, 0, 0, 0);
-    __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
+    __pyx_t_2 = PyObject_Call(__pyx_builtin_ValueError, ((PyObject *)__pyx_k_tuple_177), NULL); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3949; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+    __Pyx_GOTREF(__pyx_t_2);
+    __Pyx_Raise(__pyx_t_2, 0, 0, 0);
+    __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;
     {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3949; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
     goto __pyx_L10;
   }
@@ -18744,54 +18733,54 @@ static PyObject *__pyx_pf_6mtrand_11RandomState_92hypergeometric(struct __pyx_ob
  *             raise ValueError("ngood + nbad < nsample")
  *         return discnmN_array(self.internal_state, rk_hypergeometric, size,
  */
-  __pyx_t_3 = __Pyx_GetModuleGlobalName(__pyx_n_s__np); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3950; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
-  __Pyx_GOTREF(__pyx_t_3);
-  __pyx_t_4 = __Pyx_PyObject_GetAttrStr(__pyx_t_3, __pyx_n_s__any); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3950; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+  __pyx_t_2 = __Pyx_GetModuleGlobalName(__pyx_n_s__np); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3950; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+  __Pyx_GOTREF(__pyx_t_2);
+  __pyx_t_4 = __Pyx_PyObject_GetAttrStr(__pyx_t_2, __pyx_n_s__any); if (unlikely(!__pyx_t_4)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3950; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
   __Pyx_GOTREF(__pyx_t_4);
-  __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
-  __pyx_t_3 = __Pyx_GetModuleGlobalName(__pyx_n_s__np); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3950; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
-  __Pyx_GOTREF(__pyx_t_3);
-  __pyx_t_5 = __Pyx_PyObject_GetAttrStr(__pyx_t_3, __pyx_n_s__less); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3950; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+  __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;
+  __pyx_t_2 = __Pyx_GetModuleGlobalName(__pyx_n_s__np); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3950; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+  __Pyx_GOTREF(__pyx_t_2);
+  __pyx_t_5 = __Pyx_PyObject_GetAttrStr(__pyx_t_2, __pyx_n_s__less); if (unlikely(!__pyx_t_5)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3950; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
   __Pyx_GOTREF(__pyx_t_5);
-  __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
-  __pyx_t_3 = __Pyx_GetModuleGlobalName(__pyx_n_s__np); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3950; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
-  __Pyx_GOTREF(__pyx_t_3);
-  __pyx_t_2 = __Pyx_PyObject_GetAttrStr(__pyx_t_3, __pyx_n_s__add); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3950; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+  __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;
+  __pyx_t_2 = __Pyx_GetModuleGlobalName(__pyx_n_s__np); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3950; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
   __Pyx_GOTREF(__pyx_t_2);
-  __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
-  __pyx_t_3 = PyTuple_New(2); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3950; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+  __pyx_t_3 = __Pyx_PyObject_GetAttrStr(__pyx_t_2, __pyx_n_s__add); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3950; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
   __Pyx_GOTREF(__pyx_t_3);
+  __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;
+  __pyx_t_2 = PyTuple_New(2); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3950; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+  __Pyx_GOTREF(__pyx_t_2);
   __Pyx_INCREF(((PyObject *)__pyx_v_ongood));
-  PyTuple_SET_ITEM(__pyx_t_3, 0, ((PyObject *)__pyx_v_ongood));
+  PyTuple_SET_ITEM(__pyx_t_2, 0, ((PyObject *)__pyx_v_ongood));
   __Pyx_GIVEREF(((PyObject *)__pyx_v_ongood));
   __Pyx_INCREF(((PyObject *)__pyx_v_onbad));
-  PyTuple_SET_ITEM(__pyx_t_3, 1, ((PyObject *)__pyx_v_onbad));
+  PyTuple_SET_ITEM(__pyx_t_2, 1, ((PyObject *)__pyx_v_onbad));
   __Pyx_GIVEREF(((PyObject *)__pyx_v_onbad));
-  __pyx_t_6 = PyObject_Call(__pyx_t_2, ((PyObject *)__pyx_t_3), NULL); if (unlikely(!__pyx_t_6)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3950; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+  __pyx_t_6 = PyObject_Call(__pyx_t_3, ((PyObject *)__pyx_t_2), NULL); if (unlikely(!__pyx_t_6)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3950; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
   __Pyx_GOTREF(__pyx_t_6);
-  __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;
-  __Pyx_DECREF(((PyObject *)__pyx_t_3)); __pyx_t_3 = 0;
-  __pyx_t_3 = PyTuple_New(2); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3950; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
-  __Pyx_GOTREF(__pyx_t_3);
-  PyTuple_SET_ITEM(__pyx_t_3, 0, __pyx_t_6);
+  __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
+  __Pyx_DECREF(((PyObject *)__pyx_t_2)); __pyx_t_2 = 0;
+  __pyx_t_2 = PyTuple_New(2); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3950; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+  __Pyx_GOTREF(__pyx_t_2);
+  PyTuple_SET_ITEM(__pyx_t_2, 0, __pyx_t_6);
   __Pyx_GIVEREF(__pyx_t_6);
   __Pyx_INCREF(((PyObject *)__pyx_v_onsample));
-  PyTuple_SET_ITEM(__pyx_t_3, 1, ((PyObject *)__pyx_v_onsample));
+  PyTuple_SET_ITEM(__pyx_t_2, 1, ((PyObject *)__pyx_v_onsample));
   __Pyx_GIVEREF(((PyObject *)__pyx_v_onsample));
   __pyx_t_6 = 0;
-  __pyx_t_6 = PyObject_Call(__pyx_t_5, ((PyObject *)__pyx_t_3), NULL); if (unlikely(!__pyx_t_6)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3950; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+  __pyx_t_6 = PyObject_Call(__pyx_t_5, ((PyObject *)__pyx_t_2), NULL); if (unlikely(!__pyx_t_6)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3950; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
   __Pyx_GOTREF(__pyx_t_6);
   __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
-  __Pyx_DECREF(((PyObject *)__pyx_t_3)); __pyx_t_3 = 0;
-  __pyx_t_3 = PyTuple_New(1); if (unlikely(!__pyx_t_3)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3950; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
-  __Pyx_GOTREF(__pyx_t_3);
-  PyTuple_SET_ITEM(__pyx_t_3, 0, __pyx_t_6);
+  __Pyx_DECREF(((PyObject *)__pyx_t_2)); __pyx_t_2 = 0;
+  __pyx_t_2 = PyTuple_New(1); if (unlikely(!__pyx_t_2)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3950; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+  __Pyx_GOTREF(__pyx_t_2);
+  PyTuple_SET_ITEM(__pyx_t_2, 0, __pyx_t_6);
   __Pyx_GIVEREF(__pyx_t_6);
   __pyx_t_6 = 0;
-  __pyx_t_6 = PyObject_Call(__pyx_t_4, ((PyObject *)__pyx_t_3), NULL); if (unlikely(!__pyx_t_6)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3950; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+  __pyx_t_6 = PyObject_Call(__pyx_t_4, ((PyObject *)__pyx_t_2), NULL); if (unlikely(!__pyx_t_6)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3950; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
   __Pyx_GOTREF(__pyx_t_6);
   __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0;
-  __Pyx_DECREF(((PyObject *)__pyx_t_3)); __pyx_t_3 = 0;
+  __Pyx_DECREF(((PyObject *)__pyx_t_2)); __pyx_t_2 = 0;
   __pyx_t_1 = __Pyx_PyObject_IsTrue(__pyx_t_6); if (unlikely(__pyx_t_1 < 0)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3950; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
   __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;
   if (__pyx_t_1) {
@@ -22818,65 +22807,65 @@ static int __Pyx_InitCachedConstants(void) {
   __Pyx_GOTREF(__pyx_k_tuple_166);
   __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_166));
 
-  /* "mtrand.pyx":3928
+  /* "mtrand.pyx":3929
  *         if not PyErr_Occurred():
- *             if ngood < 1:
- *                 raise ValueError("ngood < 1")             # <<<<<<<<<<<<<<
- *             if nbad < 1:
- *                 raise ValueError("nbad < 1")
+ *             if lngood < 0:
+ *                 raise ValueError("ngood < 0")             # <<<<<<<<<<<<<<
+ *             if lnbad < 0:
+ *                 raise ValueError("nbad < 0")
  */
-  __pyx_k_tuple_168 = PyTuple_Pack(1, ((PyObject *)__pyx_kp_s_167)); if (unlikely(!__pyx_k_tuple_168)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3928; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+  __pyx_k_tuple_168 = PyTuple_Pack(1, ((PyObject *)__pyx_kp_s_167)); if (unlikely(!__pyx_k_tuple_168)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3929; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
   __Pyx_GOTREF(__pyx_k_tuple_168);
   __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_168));
 
-  /* "mtrand.pyx":3930
- *                 raise ValueError("ngood < 1")
- *             if nbad < 1:
- *                 raise ValueError("nbad < 1")             # <<<<<<<<<<<<<<
- *             if nsample < 1:
+  /* "mtrand.pyx":3931
+ *                 raise ValueError("ngood < 0")
+ *             if lnbad < 0:
+ *                 raise ValueError("nbad < 0")             # <<<<<<<<<<<<<<
+ *             if lnsample < 1:
  *                 raise ValueError("nsample < 1")
  */
-  __pyx_k_tuple_170 = PyTuple_Pack(1, ((PyObject *)__pyx_kp_s_169)); if (unlikely(!__pyx_k_tuple_170)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3930; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+  __pyx_k_tuple_170 = PyTuple_Pack(1, ((PyObject *)__pyx_kp_s_169)); if (unlikely(!__pyx_k_tuple_170)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3931; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
   __Pyx_GOTREF(__pyx_k_tuple_170);
   __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_170));
 
-  /* "mtrand.pyx":3932
- *                 raise ValueError("nbad < 1")
- *             if nsample < 1:
+  /* "mtrand.pyx":3933
+ *                 raise ValueError("nbad < 0")
+ *             if lnsample < 1:
  *                 raise ValueError("nsample < 1")             # <<<<<<<<<<<<<<
- *             if ngood + nbad < nsample:
+ *             if lngood + lnbad < lnsample:
  *                 raise ValueError("ngood + nbad < nsample")
  */
-  __pyx_k_tuple_172 = PyTuple_Pack(1, ((PyObject *)__pyx_kp_s_171)); if (unlikely(!__pyx_k_tuple_172)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3932; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+  __pyx_k_tuple_172 = PyTuple_Pack(1, ((PyObject *)__pyx_kp_s_171)); if (unlikely(!__pyx_k_tuple_172)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3933; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
   __Pyx_GOTREF(__pyx_k_tuple_172);
   __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_172));
 
-  /* "mtrand.pyx":3934
+  /* "mtrand.pyx":3935
  *                 raise ValueError("nsample < 1")
- *             if ngood + nbad < nsample:
+ *             if lngood + lnbad < lnsample:
  *                 raise ValueError("ngood + nbad < nsample")             # <<<<<<<<<<<<<<
  *             return discnmN_array_sc(self.internal_state, rk_hypergeometric, size,
  *                                     lngood, lnbad, lnsample)
  */
-  __pyx_k_tuple_174 = PyTuple_Pack(1, ((PyObject *)__pyx_kp_s_173)); if (unlikely(!__pyx_k_tuple_174)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3934; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+  __pyx_k_tuple_174 = PyTuple_Pack(1, ((PyObject *)__pyx_kp_s_173)); if (unlikely(!__pyx_k_tuple_174)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3935; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
   __Pyx_GOTREF(__pyx_k_tuple_174);
   __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_174));
 
   /* "mtrand.pyx":3945
  *         onsample = <ndarray>PyArray_FROM_OTF(nsample, NPY_LONG, NPY_ARRAY_ALIGNED)
- *         if np.any(np.less(ongood, 1)):
- *             raise ValueError("ngood < 1")             # <<<<<<<<<<<<<<
- *         if np.any(np.less(onbad, 1)):
- *             raise ValueError("nbad < 1")
+ *         if np.any(np.less(ongood, 0)):
+ *             raise ValueError("ngood < 0")             # <<<<<<<<<<<<<<
+ *         if np.any(np.less(onbad, 0)):
+ *             raise ValueError("nbad < 0")
  */
   __pyx_k_tuple_175 = PyTuple_Pack(1, ((PyObject *)__pyx_kp_s_167)); if (unlikely(!__pyx_k_tuple_175)) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 3945; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
   __Pyx_GOTREF(__pyx_k_tuple_175);
   __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_175));
 
   /* "mtrand.pyx":3947
- *             raise ValueError("ngood < 1")
- *         if np.any(np.less(onbad, 1)):
- *             raise ValueError("nbad < 1")             # <<<<<<<<<<<<<<
+ *             raise ValueError("ngood < 0")
+ *         if np.any(np.less(onbad, 0)):
+ *             raise ValueError("nbad < 0")             # <<<<<<<<<<<<<<
  *         if np.any(np.less(onsample, 1)):
  *             raise ValueError("nsample < 1")
  */
@@ -22885,7 +22874,7 @@ static int __Pyx_InitCachedConstants(void) {
   __Pyx_GIVEREF(((PyObject *)__pyx_k_tuple_176));
 
   /* "mtrand.pyx":3949
- *             raise ValueError("nbad < 1")
+ *             raise ValueError("nbad < 0")
  *         if np.any(np.less(onsample, 1)):
  *             raise ValueError("nsample < 1")             # <<<<<<<<<<<<<<
  *         if np.any(np.less(np.add(ongood, onbad),onsample)):
@@ -23132,13 +23121,13 @@ PyMODINIT_FUNC PyInit_mtrand(void)
   /*--- Variable export code ---*/
   /*--- Function export code ---*/
   /*--- Type init code ---*/
-  __pyx_ptype_6mtrand_ndarray = __Pyx_ImportType("numpy", "ndarray", sizeof(PyArrayObject), 0); if (unlikely(!__pyx_ptype_6mtrand_ndarray)) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 78; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
-  __pyx_ptype_6mtrand_broadcast = __Pyx_ImportType("numpy", "broadcast", sizeof(PyArrayMultiIterObject), 0); if (unlikely(!__pyx_ptype_6mtrand_broadcast)) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 86; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
   if (PyType_Ready(&__pyx_type_6mtrand_RandomState) < 0) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 525; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
   if (__Pyx_SetAttrString(__pyx_m, "RandomState", (PyObject *)&__pyx_type_6mtrand_RandomState) < 0) {__pyx_filename = __pyx_f[0]; __pyx_lineno = 525; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
   __pyx_ptype_6mtrand_RandomState = &__pyx_type_6mtrand_RandomState;
   __pyx_ptype_6mtrand_flatiter = __Pyx_ImportType("numpy", "flatiter", sizeof(PyArrayIterObject), 0); if (unlikely(!__pyx_ptype_6mtrand_flatiter)) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 80; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+  __pyx_ptype_6mtrand_ndarray = __Pyx_ImportType("numpy", "ndarray", sizeof(PyArrayObject), 0); if (unlikely(!__pyx_ptype_6mtrand_ndarray)) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 78; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
   __pyx_ptype_6mtrand_dtype = __Pyx_ImportType("numpy", "dtype", sizeof(PyArray_Descr), 0); if (unlikely(!__pyx_ptype_6mtrand_dtype)) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 76; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
+  __pyx_ptype_6mtrand_broadcast = __Pyx_ImportType("numpy", "broadcast", sizeof(PyArrayMultiIterObject), 0); if (unlikely(!__pyx_ptype_6mtrand_broadcast)) {__pyx_filename = __pyx_f[1]; __pyx_lineno = 86; __pyx_clineno = __LINE__; goto __pyx_L1_error;}
   /*--- Type import code ---*/
   /*--- Variable import code ---*/
   /*--- Function import code ---*/
diff --git a/numpy/random/mtrand/mtrand.pyx b/numpy/random/mtrand/mtrand.pyx
index 6e9028012..72e9a47cf 100644
--- a/numpy/random/mtrand/mtrand.pyx
+++ b/numpy/random/mtrand/mtrand.pyx
@@ -3846,20 +3846,21 @@ cdef class RandomState:
 
         Parameters
         ----------
-        ngood : float (but truncated to an integer)
-                parameter, > 0.
-        nbad  : float
-                parameter, >= 0.
-        nsample  : float
-                   parameter, > 0 and <= ngood+nbad
-        size : {tuple, int}
+        ngood : int or array_like
+            Number of ways to make a good selection.  Must be nonnegative.
+        nbad : int or array_like
+            Number of ways to make a bad selection.  Must be nonnegative.
+        nsample : int or array_like
+            Number of items sampled.  Must be at least 1 and at most
+            ``ngood + nbad``.
+        size : int or tuple of int
             Output shape.  If the given shape is, e.g., ``(m, n, k)``, then
             ``m * n * k`` samples are drawn.
 
         Returns
         -------
-        samples : {ndarray, scalar}
-                  where the values are all integers in  [0, n].
+        samples : ndarray or scalar
+            The values are all integers in  [0, n].
 
         See Also
         --------
@@ -3924,27 +3925,26 @@ cdef class RandomState:
         lnbad = PyInt_AsLong(nbad)
         lnsample = PyInt_AsLong(nsample)
         if not PyErr_Occurred():
-            if ngood < 1:
-                raise ValueError("ngood < 1")
-            if nbad < 1:
-                raise ValueError("nbad < 1")
-            if nsample < 1:
+            if lngood < 0:
+                raise ValueError("ngood < 0")
+            if lnbad < 0:
+                raise ValueError("nbad < 0")
+            if lnsample < 1:
                 raise ValueError("nsample < 1")
-            if ngood + nbad < nsample:
+            if lngood + lnbad < lnsample:
                 raise ValueError("ngood + nbad < nsample")
             return discnmN_array_sc(self.internal_state, rk_hypergeometric, size,
                                     lngood, lnbad, lnsample)
 
-
         PyErr_Clear()
 
         ongood = <ndarray>PyArray_FROM_OTF(ngood, NPY_LONG, NPY_ARRAY_ALIGNED)
         onbad = <ndarray>PyArray_FROM_OTF(nbad, NPY_LONG, NPY_ARRAY_ALIGNED)
         onsample = <ndarray>PyArray_FROM_OTF(nsample, NPY_LONG, NPY_ARRAY_ALIGNED)
-        if np.any(np.less(ongood, 1)):
-            raise ValueError("ngood < 1")
-        if np.any(np.less(onbad, 1)):
-            raise ValueError("nbad < 1")
+        if np.any(np.less(ongood, 0)):
+            raise ValueError("ngood < 0")
+        if np.any(np.less(onbad, 0)):
+            raise ValueError("nbad < 0")
         if np.any(np.less(onsample, 1)):
             raise ValueError("nsample < 1")
         if np.any(np.less(np.add(ongood, onbad),onsample)):
diff --git a/numpy/random/tests/test_random.py b/numpy/random/tests/test_random.py
index c9ced3569..f49592f7d 100644
--- a/numpy/random/tests/test_random.py
+++ b/numpy/random/tests/test_random.py
@@ -303,6 +303,24 @@ class TestRandomDist(TestCase):
                          [ 9,  9]])
         np.testing.assert_array_equal(actual, desired)
 
+        # Test nbad = 0
+        actual = np.random.hypergeometric(5, 0, 3, size=4)
+        desired = np.array([3, 3, 3, 3])
+        np.testing.assert_array_equal(actual, desired)
+
+        actual = np.random.hypergeometric(15, 0, 12, size=4)
+        desired = np.array([12, 12, 12, 12])
+        np.testing.assert_array_equal(actual, desired)
+
+        # Test ngood = 0
+        actual = np.random.hypergeometric(0, 5, 3, size=4)
+        desired = np.array([0, 0, 0, 0])
+        np.testing.assert_array_equal(actual, desired)
+
+        actual = np.random.hypergeometric(0, 15, 12, size=4)
+        desired = np.array([0, 0, 0, 0])
+        np.testing.assert_array_equal(actual, desired)
+
     def test_laplace(self):
         np.random.seed(self.seed)
         actual = np.random.laplace(loc=.123456789, scale=2.0, size=(3, 2))