diff --git a/scikits/learn/datasets/narr/__init__.py b/scikits/learn/datasets/narr/__init__.py
new file mode 100644
index 0000000000000000000000000000000000000000..a60cd86263d50435f23fd02a220d57ec1f02b9e1
--- /dev/null
+++ b/scikits/learn/datasets/narr/__init__.py
@@ -0,0 +1,8 @@
+#! /usr/bin/env python
+# Last Change: Fri Jun 08 12:00 PM 2007 J
+import data as _narr
+__doc__ = _narr.DESCRLONG
+copyright = _narr.COPYRIGHT
+source = _narr.SOURCE
+
+load = _narr.load
diff --git a/scikits/learn/datasets/narr/data.py b/scikits/learn/datasets/narr/data.py
new file mode 100644
index 0000000000000000000000000000000000000000..d2a5e216ad5a6a38325af67457ca280d580fd388
--- /dev/null
+++ b/scikits/learn/datasets/narr/data.py
@@ -0,0 +1,182 @@
+"""
+=============
+NCEP NARR data sets
+=============
+
+This module provides access to the National Center for Environment Prediction
+(NCEP) North American Regional Reanalysis (NARR) data sets through the
+Data Access Protocol (DAP). The documentation describing the data set
+variables is included in the doc directory.
+
+See the `pydap home page`_ for information on the DAP client.
+
+.. _pydap home page: http://pydap.org/
+"""
+
+__docformat__ = 'restructuredtext'
+
+TITLE = "NCEP North American Regional Reanalysis (NARR)"
+
+COPYRIGHT = """The information on government servers are in the public domain,
+unless specifically annotated otherwise, and may be used freely by the public so
+long as you do not 1) claim it is your own (e.g. by claiming copyright for NWS
+information -- see below), 2) use it in a manner that implies an endorsement
+or affiliation with NOAA/NWS, or 3) modify it in content and then present it as
+official government material. You also cannot present information of your own
+in a way that makes it appear to be official government information..
+
+The user assumes the entire risk related to its use of this data. NWS is
+providing this data "as is," and NWS disclaims any and all warranties, whether
+express or implied, including (without limitation) any implied warranties of
+merchantability or fitness for a particular purpose. In no event will NWS be
+liable to you or to any third party for any direct, indirect, incidental,
+consequential, special or exemplary damages or lost profit resulting from any
+use or misuse of this data.
+
+As required by 17 U.S.C. 403, third parties producing copyrighted works
+consisting predominantly of the material appearing in NWS Web pages must
+provide notice with such work(s) identifying the NWS material incorporated
+and stating that such material is not subject to copyright protection."""
+
+DISCLAIMER = """While every effort has been made to ensure that data are
+accurate and reliable within the limits of the current state of
+the art, NOAA cannot assume liability for any damages caused by
+any errors or omissions in the data, nor as a result of the
+failure of the data to function on a particular system.
+NOAA makes no warranty, expressed or implied, nor does the fact
+of distribution constitute such a warranty.
+The data used to produce these analyses and forecasts has
+undergone automated quality checks.
+"""
+
+SOURCE = """National Centers for Environmental Prediction (NCEP) _`North American Regional Reanalysis (NARR)` _`NOMADS web interface`.
+
+.. __North American Regional Reanalysis (NARR): http://wwwt.emc.ncep.noaa.gov/mmb/rreanl/index.html
+.. _NOMADS web interface: http://nomads.ncdc.noaa.gov/data.php#narr_datasets
+
+References:
+ Mesinger, F., et al, 2004: NCEP North American Regional
+Reanalysis, 15th Symp. On Global Change and Climate Variations,
+Seattle, WA, 11-15 Jan 2004.
+
+ Shafran, P., J. Woollen, W. Ebisuzaki, W. Shi, Y. Fan, R. W.
+Grumbine, M. Fennessy, 2004: Observational Data Used for Assimilation in
+the NCEP North American Regional Reanalysis, 20th Intl. Conf. On Interactive
+Information Processing Systems for Meteor. Ocean. And Hydrology. Seattle, WA, 11-15
+Jan 2004.
+
+ Ebisuzaki, W., J. Alpert, J. Wang, D. Jovic, P. Shafran, 2004: North American
+Regional Reanalysis: end user access to large data sets, 20th Intl. Conf. On
+Interactive Information Processing Systems for Meteor. Ocean. And Hydrology. Seattle,
+WA, 11-15 Jan 2004.
+
+ Mesinger, F., G. DiMego, E. Kalnay, P. Shafran, W. Ebisuzaki, Y. Fan, R. Grumbine,
+W. Higgins, Y. Lin, K. Mitchell, D. Parrish, E. Rogers, W. Shi, D. Stokes, J. Woolen,
+2003: NCEP Regional Reanalysis, Symp. on Observing and Understanding the Variability
+of Water in Weather and Climate, Long Beach, CA, Feb.9-13, 2003.
+"""
+
+DESCRSHORT = """NCEP NARR data sets"""
+
+DESCRLONG = """The NARR is a reanalysis of
+historical observations using a 32 km version of the NCEP 1993
+operational ETA model and ETA data assimilation system (EDAS).
+The domain of analyses includes North and Central America as
+well as small parts of the UK, Eastern Asia and South America
+and the oceans in between. The period of the reanalyses is from
+October 1978 to the present and analyses were made 8 times
+daily. Horizontal boundary conditions were derived from the
+NCEP/DOE Reanalysis.
+The "merged" dataset provides a high spatial (32 km) and
+temporal (3 hour) analyses of North America and adjacent oceans
+and land masses from October 1978 to the present. Advantages
+over the widely used NCEP/NCAR Reanalysis are its higher
+resolution and a much better treatment of the land surface
+through a better land-surface model (NOAH), through the
+assimilation of more surface data (observed precipitation and
+surface winds) and through a better representation of the
+terrain (heights, vegetation, soil type).
+This data set contains "conventional" atmospheric analyses as
+well as model-derived fields which contain estimates of
+subsurface, surface, and radiative properties.
+"""
+
+import dap.client
+import datetime
+import numpy as np
+
+NARR_A_URL = "http://nomads.ncdc.noaa.gov:9091/dods/NCEP_NARR_DAILY"
+
+def load(year, month, day, verbose=True):
+ """
+ Load the National Center for Environment Prediction (NCEP) North American
+ Regional Reanalysis (NARR) data at the given date. The data is available
+ through Data Access Protocol (DAP) from a remote server. The returned
+ object is not the entire data set, but only the metadata describing the
+ data set content. Individual variables are properties of the object, and
+ are downloaded only on explicit request (see example below).
+
+ Parameters:
+ year : int
+ month : int
+ day : int
+
+ Note:
+ The data is available starting on 1979-01-01 and up to the the month before
+ last. Each dataset contains around 190 different variables at 3h intervals.
+ The DAP protocol was formerly known as the Distributed Oceanographic Data
+ Systems (DODS).
+
+
+ Some variables included in the set:
+ ugrd10m : Wind speed (u) at 10 m altitude [m/s].
+ vgrd10m : Wind speed (v) at 10 m altitude [m/s].
+ mslet: Mean sea level pressure (eta model) [pa].
+ tmpsfc : Surface temperature [K].
+ dswrfsfc : Surface downward shortwave radiation flux [W/m^2].
+ time : Days since 1-1-1 00:00:00.
+ lat : Latitude (0 to 89.625) at 0.375 degree resolution [deg].
+ lon : Longitude (-220 to -0.625) at 0.375 degree resolution [deg].
+
+ Example:
+ narr = load(2005, 6, 23)
+ temp = narr.tmpsfc[:][:][:] # Slices refer to time, lat, lon.
+ time = narr.time[:]
+ lat = narr.lat[:]
+ lon = narr.lon[:]
+
+ """
+
+ tail = 'narr-a_221_'
+ dir1 = '%d%02d'%(year, month)
+ dir2 = dir1+'%02d'%day
+ tail = tail +dir2+'_0000_000'
+ url = '/'.join([NARR_A_URL, dir1, dir2, tail])
+ if verbose:
+ print 'Accessing DODS server: ', url
+ return dap.client.open(url)
+
+
+
+def convert_time(time):
+ """Return datetime objects.
+
+ The NARR dataset provides time in days since 1-1-1 00:00:00:0.0. This
+ function converts the NARR time into a datetime.datetime object.
+
+ Note:
+ There seems to be a discrepancy between the days counting of NARR and
+ python datetime. To achieve the fit, two days must be substracted from the
+ NARR time.
+ """
+ time = time-2
+ t0 = datetime.datetime(1,1, 1, 0,0,0)
+ hour = datetime.timedelta(hours=1)
+ day = datetime.timedelta(days=1)
+ time = np.atleast_1d(time)
+ days = time.astype(int)
+ hours = ((time%1)*24).astype(int)
+ return t0+days*day+hours*hour
+
+
+narr = load(1979, 1,1)
diff --git a/scikits/learn/datasets/narr/doc/narr_data_documentation.pdf b/scikits/learn/datasets/narr/doc/narr_data_documentation.pdf
new file mode 100644
index 0000000000000000000000000000000000000000..fe5727b11e94a1d230f131a75a66c067db7da588
Binary files /dev/null and b/scikits/learn/datasets/narr/doc/narr_data_documentation.pdf differ
diff --git a/scikits/learn/datasets/narr/doc/overview.pdf b/scikits/learn/datasets/narr/doc/overview.pdf
new file mode 100644
index 0000000000000000000000000000000000000000..aedba53e76d2743195e14ae9746846bdb5d4766a
Binary files /dev/null and b/scikits/learn/datasets/narr/doc/overview.pdf differ
diff --git a/scikits/learn/datasets/setup.py b/scikits/learn/datasets/setup.py
index 4926774d3931a39a1827eda8e271a6c2e8b4a7c9..e7493dff0f03bd5180deae7464da71ec943983e5 100644
--- a/scikits/learn/datasets/setup.py
+++ b/scikits/learn/datasets/setup.py
@@ -8,6 +8,8 @@ def configuration(parent_package='',top_path=None):
config.add_subpackage('iris')
config.add_subpackage('heart')
config.add_subpackage('german')
+ config.add_subpackage('nwis')
+ config.add_subpackage('narr')
config.add_subpackage('misc')
config.make_config_py() # installs __config__.py
return config