diff --git a/.travis.yml b/.travis.yml
index b7146cd8da..17c7a60b50 100644
--- a/.travis.yml
+++ b/.travis.yml
@@ -93,7 +93,7 @@ script:
   - set -e
   - pytest --cov-config .coveragerc --cov=arch arch --durations=10
   - python ci/performance.py
-  - if [[ ${DOCBUILD} = true ]]; then cd doc && make html && cd .. ; fi
+  - if [[ ${DOCBUILD} = true ]]; then python setup.py install && cd doc && make html && cd .. ; fi
   - if [[ ${DOCBUILD} = true && ${TRAVIS_BRANCH} = "master" ]]; then doctr deploy doc; fi
   - flake8 arch
 
diff --git a/arch/unitroot/unitroot.py b/arch/unitroot/unitroot.py
index c5bb8e3715..7b2a7ea76a 100644
--- a/arch/unitroot/unitroot.py
+++ b/arch/unitroot/unitroot.py
@@ -45,9 +45,9 @@ def _autolag_ols(endog, exog, startlag, maxlag, method, modargs=(), regresults=F
 
     Parameters
     ----------
-    endog : array-like
+    endog : {ndarray, Series}
         nobs array containing endogenous variable
-    exog : array-like
+    exog : {ndarray, DataFrame}
         nobs by (startlag + maxlag) array containing lags and possibly other
         variables
     startlag : int
@@ -123,7 +123,7 @@ def _df_select_lags(y, trend, max_lags, method):
 
     Parameters
     ----------
-    y : array
+    y : ndarray
         The data for the lag selection exercise
     trend : {'nc','c','ct','ctt'}
         The trend order
@@ -172,7 +172,7 @@ def _estimate_df_regression(y, trend, lags):
 
     Parameters
     ----------
-    y : array
+    y : ndarray
         The data for the lag selection
     trend : {'nc','c','ct','ctt'}
         The trend order
@@ -384,7 +384,7 @@ class ADF(UnitRootTest):
 
     Parameters
     ----------
-    y : array
+    y : {ndarray, Series}
         The data to test for a unit root
     lags : int, optional
         The number of lags to use in the ADF regression.  If omitted or None,
@@ -529,7 +529,7 @@ class DFGLS(UnitRootTest):
 
     Parameters
     ----------
-    y : array
+    y : {ndarray, Series}
         The data to test for a unit root
     lags : int, optional
         The number of lags to use in the ADF regression.  If omitted or None,
@@ -692,7 +692,7 @@ class PhillipsPerron(UnitRootTest):
 
     Parameters
     ----------
-    y : array
+    y : {ndarray, Series}
         The data to test for a unit root
     lags : int, optional
         The number of lags to use in the Newey-West estimator of the long-run
@@ -870,7 +870,7 @@ class KPSS(UnitRootTest):
 
     Parameters
     ----------
-    y : array
+    y : {ndarray, Series}
         The data to test for stationarity
     lags : int, optional
         The number of lags to use in the Newey-West estimator of the long-run
@@ -963,7 +963,7 @@ class VarianceRatio(UnitRootTest):
 
     Parameters
     ----------
-    y : array
+    y : {ndarray, Series}
         The data to test for a random walk
     lags : int
         The number of periods to used in the multi-period variance, which is
@@ -1252,7 +1252,7 @@ def mackinnoncrit(num_unit_roots=1, regression="c", nobs=inf,
 
     Returns
     -------
-    crit_vals : array
+    crit_vals : ndarray
         Three critical values corresponding to 1%, 5% and 10% cut-offs.
 
     Notes
@@ -1311,7 +1311,7 @@ def kpss_crit(stat, trend='c'):
     -------
     pvalue : float
         The interpolated p-value
-    crit_val : array
+    crit_val : ndarray
         Three element array containing the 10%, 5% and 1% critical values,
         in order
 
diff --git a/arch/univariate/base.py b/arch/univariate/base.py
index 0a1f58850d..fe13e3a47b 100644
--- a/arch/univariate/base.py
+++ b/arch/univariate/base.py
@@ -37,7 +37,7 @@ def _callback(*args):
 
     Parameters
     ----------
-    parameters : np.ndarray
+    parameters : : ndarray
         Parameter value (not used by function)
 
     Notes
@@ -60,9 +60,9 @@ def constraint(a, b):
 
     Parameters
     ----------
-    a : np.ndarray
+    a : ndarray
         Parameter loadings
-    b : np.ndarray
+    b : ndarray
         Constraint bounds
 
     Returns
@@ -113,7 +113,7 @@ def implicit_constant(x):
 
     Parameters
     ----------
-    x : np.ndarray
+    x : ndarray
         Array to be tested
 
     Returns
@@ -178,9 +178,9 @@ def constraints(self):
 
         Returns
         -------
-        a : np.ndarray
+        a : ndarray
             Number of constraints by number of parameters loading array
-        b : np.ndarray
+        b : ndarray
             Number of constraints array of lower bounds
 
         Notes
@@ -387,7 +387,7 @@ def fit(self, update_freq=1, disp='final', starting_values=None,
         disp : str
             Either 'final' to print optimization result or 'off' to display
             nothing
-        starting_values : np.ndarray, optional
+        starting_values : ndarray, optional
             Array of starting values to use.  If not provided, starting values
             are constructed by the model components.
         cov_type : str, optional
@@ -572,7 +572,7 @@ def starting_values(self):
 
         Returns
         -------
-        sv : np.ndarray
+        sv : ndarray
             Starting values
         """
         params = np.asarray(self._fit_no_arch_normal_errors().params)
@@ -599,16 +599,16 @@ def resids(self, params, y=None, regressors=None):
 
         Parameters
         ----------
-        params : np.ndarray
+        params : ndarray
             Model parameters
-        y : np.ndarray, optional
+        y : ndarray, optional
             Alternative values to use when computing model residuals
-        regressors : np.ndarray, optional
+        regressors : ndarray, optional
             Alternative regressor values to use when computing model residuals
 
         Returns
         -------
-        resids : np.ndarray
+        resids : ndarray
             Model residuals
         """
         raise NotImplementedError('Subclasses must implement')
@@ -619,7 +619,7 @@ def compute_param_cov(self, params, backcast=None, robust=True):
 
         Parameters
         ----------
-        params : np.ndarray
+        params : ndarray
             Model parameters
         backcast : float
             Value to use for pre-sample observations
@@ -660,7 +660,7 @@ def forecast(self, params, horizon=1, start=None, align='origin', method='analyt
 
         Parameters
         ----------
-        params : np.ndarray, optional
+        params : ndarray, optional
             Alternative parameters to use.  If not provided, the parameters
             estimated when fitting the model are used.  Must be identical in
             shape to the parameters computed by fitting the model.
@@ -751,12 +751,12 @@ class ARCHModelFixedResult(_SummaryRepr):
 
     Parameters
     ----------
-    params : np.ndarray
+    params : ndarray
         Estimated parameters
-    resid : np.ndarray
+    resid : ndarray
         Residuals from model.  Residuals have same shape as original data and
         contain nan-values in locations not used in estimation
-    volatility : np.ndarray
+    volatility : ndarray
         Conditional volatility from model
     dep_var: Series
         Dependent variable
@@ -786,7 +786,7 @@ class ARCHModelFixedResult(_SummaryRepr):
         Akaike information criteria
     bic : float
         Schwarz/Bayes information criteria
-    conditional_volatility : {np.ndarray, Series}
+    conditional_volatility : {ndarray, Series}
         nobs element array containing the conditional volatility (square root
         of conditional variance).  The values are aligned with the input data
         so that the value in the t-th position is the variance of t-th error,
@@ -797,7 +797,7 @@ class ARCHModelFixedResult(_SummaryRepr):
         Number of observations used in the estimation
     num_params : int
         Number of parameters in the model
-    resid : {np.ndarray, Series}
+    resid : {ndarray, Series}
         nobs element array containing model residuals
     model : ARCHModel
         Model instance used to produce the fit
@@ -1056,7 +1056,7 @@ def forecast(self, params=None, horizon=1, start=None, align='origin', method='a
 
         Parameters
         ----------
-        params : np.ndarray, optional
+        params : ndarray, optional
             Alternative parameters to use.  If not provided, the parameters
             estimated when fitting the model are used.  Must be identical in
             shape to the parameters computed by fitting the model.
@@ -1126,7 +1126,7 @@ def hedgehog_plot(self, params=None, horizon=10, step=10, start=None,
 
         Parameters
         ----------
-        params : {np.ndarray, Series}
+        params : {ndarray, Series}
             Alternative parameters to use.  If not provided, the parameters
             computed by fitting the model are used.  Must be 1-d and identical
             in shape to the parameters computed by fitting the model.
@@ -1225,18 +1225,18 @@ class ARCHModelResult(ARCHModelFixedResult):
 
     Parameters
     ----------
-    params : np.ndarray
+    params : ndarray
         Estimated parameters
-    param_cov : {np.ndarray, None}
+    param_cov : {ndarray, None}
         Estimated variance-covariance matrix of params.  If none, calls method
         to compute variance from model when parameter covariance is first used
         from result
     r2 : float
         Model R-squared
-    resid : np.ndarray
+    resid : ndarray
         Residuals from model.  Residuals have same shape as original data and
         contain nan-values in locations not used in estimation
-    volatility : np.ndarray
+    volatility : ndarray
         Conditional volatility from model
     cov_type : str
         String describing the covariance estimator used
@@ -1273,7 +1273,7 @@ class ARCHModelResult(ARCHModelFixedResult):
         Akaike information criteria
     bic : float
         Schwarz/Bayes information criteria
-    conditional_volatility : {np.ndarray, Series}
+    conditional_volatility : {ndarray, Series}
         nobs element array containing the conditional volatility (square root
         of conditional variance).  The values are aligned with the input data
         so that the value in the t-th position is the variance of t-th error,
@@ -1296,7 +1296,7 @@ class ARCHModelResult(ARCHModelFixedResult):
         Array of parameter standard errors
     pvalues : Series
         Array of p-values for the t-statistics
-    resid : {np.ndarray, Series}
+    resid : {ndarray, Series}
         nobs element array containing model residuals
     model : ARCHModel
         Model instance used to produce the fit
@@ -1324,7 +1324,7 @@ def conf_int(self, alpha=0.05):
 
         Returns
         -------
-        ci : np.ndarray
+        ci : ndarray
             Array where the ith row contains the confidence interval  for the
             ith parameter
         """
@@ -1598,14 +1598,14 @@ class ARCHModelForecast(object):
 
     Parameters
     ----------
-    index : {list, np.ndarray}
-    mean : np.ndarray
-    variance : np.ndarray
-    residual_variance : np.ndarray
-    simulated_paths : np.ndarray, optional
-    simulated_variances : np.ndarray, optional
-    simulated_residual_variances : np.ndarray, optional
-    simulated_residuals : np.ndarray, optional
+    index : {list, ndarray}
+    mean : ndarray
+    variance : ndarray
+    residual_variance : ndarray
+    simulated_paths : ndarray, optional
+    simulated_variances : ndarray, optional
+    simulated_residual_variances : ndarray, optional
+    simulated_residuals : ndarray, optional
     align : {'origin', 'target'}
 
     Attributes
diff --git a/arch/univariate/mean.py b/arch/univariate/mean.py
index 466274d3e1..077e122706 100644
--- a/arch/univariate/mean.py
+++ b/arch/univariate/mean.py
@@ -10,6 +10,7 @@
 import numpy as np
 from numpy import zeros, empty, ones, isscalar, log
 from pandas import DataFrame
+import pandas as pd
 from scipy.optimize import OptimizeResult
 from statsmodels.tools.decorators import cache_readonly
 from statsmodels.tsa.tsatools import lagmat
@@ -41,11 +42,11 @@ def _ar_forecast(y, horizon, start_index, constant, arp):
 
     Parameters
     ----------
-    y : np.ndarray
+    y : ndarray
     horizon : int
     start_index : int
     constant : float
-    arp : np.ndarray
+    arp : ndarray
 
     Returns
     -------
@@ -85,9 +86,9 @@ class HARX(ARCHModel):
 
     Parameters
     ----------
-    y : {np.ndarray, Series}
+    y : {ndarray, Series}
         nobs element vector containing the dependent variable
-    x : {np.ndarrayor, DataFrame}, optional
+    x : {ndarrayor, DataFrame}, optional
         nobs by k element array containing exogenous regressors
     lags : {scalar, array}, optional
         Description of lag structure of the HAR.  Scalar included all lags
@@ -259,7 +260,7 @@ def simulate(self, params, nobs, burn=500, initial_value=None, x=None,
 
         Parameters
         ----------
-        params : np.ndarray
+        params : ndarray
             Parameters to use when simulating the model.  Parameter order is
             [mean volatility distribution] where the parameters of the mean
             model are ordered [constant lag[0] lag[1] ... lag[p] ex[0] ...
@@ -271,13 +272,13 @@ def simulate(self, params, nobs, burn=500, initial_value=None, x=None,
         burn : int, optional
             Number of values to simulate to initialize the model and remove
             dependence on initial values.
-        initial_value : {np.ndarray, float}, optional
+        initial_value : {ndarray, float}, optional
             Either a scalar value or `max(lags)` array set of initial values to
             use when initializing the model.  If omitted, 0.0 is used.
         x : array, optional
             nobs + burn by k array of exogenous variables to include in the
             simulation.
-        initial_value_vol : {np.ndarray, float}, optional
+        initial_value_vol : {ndarray, float}, optional
             An array or scalar to use when initializing the volatility process.
 
         Returns
@@ -705,7 +706,7 @@ class ConstantMean(HARX):
 
     Parameters
     ----------
-    y : array or Series
+    y : {ndarray, Series}
         nobs element vector containing the dependent variable
     hold_back : int
         Number of observations at the start of the sample to exclude when
@@ -813,7 +814,7 @@ class ZeroMean(HARX):
 
     Parameters
     ----------
-    y : array or Series
+    y : {ndarray, Series}
         nobs element vector containing the dependent variable
     hold_back : int
         Number of observations at the start of the sample to exclude when
@@ -927,9 +928,9 @@ class ARX(HARX):
 
     Parameters
     ----------
-    y : array or Series
+    y : {ndarray, Series}
         nobs element vector containing the dependent variable
-    x : array or DataFrame, optional
+    x : {ndarray, DataFrame}, optional
         nobs by k element array containing exogenous regressors
     lags : scalar, 1-d array, optional
         Description of lag structure of the HAR.  Scalar included all lags
@@ -1027,9 +1028,9 @@ class LS(HARX):
 
     Parameters
     ----------
-    y : array or Series
+    y : {ndarray, Series}
         nobs element vector containing the dependent variable
-    x : array or DataFrame, optional
+    y : {ndarray, DataFrame}, optional
         nobs by k element array containing exogenous regressors
     constant : bool, optional
         Flag whether the model should include a constant
@@ -1073,7 +1074,7 @@ def arch_model(y, x=None, mean='Constant', lags=0, vol='Garch', p=1, o=0, q=1,
 
     Parameters
     ----------
-    y : {np.ndarray, Series, None}
+    y : {ndarray, Series, None}
         The dependent variable
     x : {np.array, DataFrame}, optional
         Exogenous regressors.  Ignored if model does not permit exogenous
diff --git a/arch/univariate/volatility.py b/arch/univariate/volatility.py
index 2cefdaaa1f..6c639c2913 100644
--- a/arch/univariate/volatility.py
+++ b/arch/univariate/volatility.py
@@ -1922,7 +1922,7 @@ class FixedVariance(VolatilityProcess):
 
     Parameters
     ----------
-    variance : {array, pd.Series}
+    variance : {array, Series}
         Array containing the variances to use.  Shoule have the same shape as the data used in the
         model.
     unit_scale : bool, optional
diff --git a/arch/utility/array.py b/arch/utility/array.py
index a6f410e396..75e0d61f66 100644
--- a/arch/utility/array.py
+++ b/arch/utility/array.py
@@ -122,7 +122,7 @@ def date_to_index(date, date_index):
     ----------
     date : string, datetime or datetime64
         Date to use when returning the index
-    date_index : array
+    date_index : {list, ndarray}
         Index data containing datetime64 values
 
     Returns
diff --git a/arch/utility/cov.py b/arch/utility/cov.py
index 4b5f45228d..9a47951946 100644
--- a/arch/utility/cov.py
+++ b/arch/utility/cov.py
@@ -7,7 +7,7 @@ def cov_nw(y, lags=0, demean=True, axis=0, ddof=0):
 
     Parameters
     ----------
-    y : array
+    y : ndarray
         Values to use when computing the Newey-West covariance estimator, either
         1-d or 2-d. When y is 2d, default behavior is to treat columns as variables
         and rows as observations.
@@ -23,7 +23,7 @@ def cov_nw(y, lags=0, demean=True, axis=0, ddof=0):
 
     Returns
     -------
-    cov : array
+    cov : ndarray
         The estimated covariance
 
     """
diff --git a/arch/utility/timeseries.py b/arch/utility/timeseries.py
index 9b8b256a3e..f853077d9b 100644
--- a/arch/utility/timeseries.py
+++ b/arch/utility/timeseries.py
@@ -43,7 +43,7 @@ def add_trend(x=None, trend="c", prepend=False, nobs=None,
 
     Parameters
     ----------
-    x : array-like or None
+    x : {ndarray, DataFrame}
         Original array of data. If None, then nobs must be a positive integer
     trend : str {"c","t","ct","ctt"}
         "c" add constant only
diff --git a/doc/source/conf.py b/doc/source/conf.py
index 24a70ee775..bd4c9ae7fd 100644
--- a/doc/source/conf.py
+++ b/doc/source/conf.py
@@ -39,6 +39,7 @@
     'sphinx.ext.coverage',
     'sphinx.ext.ifconfig',
     'sphinx.ext.napoleon',
+    'sphinx.ext.intersphinx',
     'IPython.sphinxext.ipython_console_highlighting',
     'IPython.sphinxext.ipython_directive'
 ]
@@ -290,11 +291,13 @@
 # Example configuration for intersphinx: refer to the Python standard library.
 intersphinx_mapping = {
     'statsmodels': ('http://www.statsmodels.org/dev/', None),
-    'matplotlib': ('http://matplotlib.org/', None),
-    'scipy': ('http://docs.scipy.org/doc/scipy/reference/', None),
-    'python': ('http://docs.python.org/3', None),
-    'numpy': ('http://docs.scipy.org/doc/numpy', None),
-    'pandas': ('http://pandas.pydata.org/pandas-docs/stable/', None)
+    'matplotlib': ('https://matplotlib.org', None),
+    'scipy': ('https://docs.scipy.org/doc/scipy/reference/', None),
+    'python': ('https://docs.python.org/3', None),
+    'numpy': ('https://docs.scipy.org/doc/numpy', None),
+    'np': ('https://docs.scipy.org/doc/numpy', None),
+    'pandas': ('https://pandas.pydata.org/pandas-docs/stable/', None),
+    'pd': ('https://pandas.pydata.org/pandas-docs/stable/', None),
 }
 
 extlinks = {'issue': ('https://github.com/bashtage/arch/issues/%s', 'GH')}