Scripts for customized regridding of Level-2 data to Level-3 data
Project description
=================================================================
Wisconsin Horizontal Interpolation Program for Satellites (WHIPS)
=================================================================
**Project Title:** WHIPS
**Purpose of Project:** Provide a well-documented, easy-to-use general-purpose
processing module for processing satellite data
**Version:** 0.2.82 (2/3/2012)
**Authors:** oberman, maki, strom
**Contact:** oberman@wisc.edu
QUICK START
===========
1. Install the program and run the built-in test module to confirm
that it is working properly. Installation instructions can be found
in the file INSTALL.txt
2. Download whatever data you plan to process. Currently, the program
is designed to process OMI NO2 DOMINO level 2 data, OMI NO2 NASA level
2 data and MOPITT CO data. See the detailed documentation for the
--filelist argument for locations of data.
3. Navigate to the folder where process.py is located (or add it to
your path) and invoke it as:
process.py --help
4. Follow the on-screen instructions, adding each of the required
parameters. If you need help with the projection attributes or the
output function attributes, invoke the built-in help as:
process.py --AttributeHelp <function_name>
For detailed explanations of all parameters and attributes, see the
"Parameter Details" section below.
5. Invoke process.py once for each output file you'd like to create.
Note that the software creates output files with only a single
timestep, so you'll need to invoke the command once for each timestep
(IE if you want a month with timesteps every day, you'll probably want
to write a shell script that calls the command once for each day)
6. Additionally, you may create a grid file for your chosen grid by
including the --includeGrid flag followed by the desired filename.
A file containing the gridcells used by the projection will be written
to the same directory as the standard output file.
7. Concatenate your outputs if desired (the authors recommend the NCO
operators at http://nco.sourceforge.net/ if you're using a netCDF
output format) and carry on!
EXAMPLE/TEMPLATE INVOCATIONS
============================
For clarity and readability, line continuation characters are used to
place each attribute on a separate line. It is not required to break
up attributes like this, but the command line needs to see the
invocation as a single command, so if you want to break it onto
multple lines you must use line-continuation characters.
1. Process MOPITT level 2 CO data, (Version 5)
----------------------------------------------
- processes a single file
- Uses a 36km lambert conic conformal grid centered over North
America
- Writes out a 2D, 3D, and 4D parameter from the file
process.py \
--directory /path/to/input/files/ \
--filetype HDFmopittl2 \
--fileList MOP02T-20050106-L2V10.1.1.prov.hdf \
--gridProj lcc2par \
--projAttrs xOrig:-2916000 yCell:36000 refLon:-97 refLat:40 \
nCols:162 nRows:126 stdPar2:45 stdPar1:33 xCell:36000 \
earthRadius:6370000 yOrig:-2268000\
--mapFunc point_in_cell \
--outFunc unweighted_filtered_MOPITT_avg_netCDF \
--outFuncAttrs time:Time longitude:Longitude \
"inFieldNames:Time,Retrieved CO Mixing Ratio Profile,Retrieved CO Surface Mixing Ratio" \
outFieldNames:time,COprof,COsurf outUnits:TAI93,ppbv,ppbv \
logNormal:False,True,True "dimLabels:;layer,valOrStdDev;valOrStdDev" \
"dimSizes:;9,2;2" "timeStart:00:00:00_01-06-2005" \
"timeStop:23:59:59_01-06-2005" timeComparison:UTC \
fillVal:-9999.0 \
solZenAngCutoff:85 \
"solZenAng:Solar Zenith Angle" dayTime:True \
"surfTypeField:Surface Index" \
"colMeasField:Retrieved CO Mixing Ratio Profile" \
--outDirectory /path/to/output/directory/ \
--outFileName MOPITT_v5_20050106_daytime_CONUS36km_test.nc \
--verbose True \
--interactive True
2. Process OMI level 2 DOMINO NO2 data, (Version 5)
---------------------------------------------------
- processes any number of files
- Uses a 36km lambert conic conformal grid centered over North
America
- Writes out a 2D, and 3D parameter from the file
process.py \
--directory /where/you/keep/the/files/ \
--filetype HDFknmiomil2 \
--fileList OMI-Aura_L2-OMDOMINO_2006m0701t0023-o10423_v003-2010m1008t224420.he5 \
OMI-Aura_L2-OMDOMINO_2006m0701t0112-o10424_v003-20120m1008t224845.he5 \
--gridProj lcc2par \
--projAttrs xOrig:-2916000 yCell:36000 refLon:-97 refLat:40 \
nCols:162 nRows:126 stdPar2:45 stdPar1:33 xCell:36000 \
earthRadius:6370000 yOrig:-2268000 \
--mapFunc regional_intersect \
--outFunc OMNO2e_netCDF_avg \
--outFuncAttrs overallQualFlag:TropColumnFlag "cloudFrac:Cloud Fraction" \
solarZenithAngle:SolarZenithAngle time:Time longitude:Longitude \
inFieldNames:Time,AveragingKernel,TroposphericVerticalColumn \
outFieldNames:time,avKern,tropVCD \
"outUnits:TAI93,unitless x 1000,molec/cm^2x1^-15" extraDimLabel:none,Layers,none \
extraDimSize:0,34,0 timeStart:00:00:00_07-01-2006 timeStop:23:59:59_07-01-2006 \
timeComparison:UTC fillVal:-9999 cloudFractUpperCutoff:0.3 \
solarZenAngUpperCutoff:85 pixIndXtrackAxis:1 \
--outDirectory /where/you/want/output \
--outFileName OMI_DOMINO_20060701_test.nc \
--includeGrid OMI_DOMINO_GridFileName.nc \
--verbose True \
--interactive True
PARAMETER DETAILS
=================
Each input attribute is explained here. Note that each output
function has a separate set of required inputs and that the
--outFuncAttrs is therefore broken down by output function. Make sure
you're referencing the details for the attributes relevant to the
function you want to use.
* --help
REQUIRED: NO
DEFAULT: N/A
- Display the onscreen help message and exit the program.
* --directory /path/to/input/directory
REQUIRED: NO
DEFAULT: the current working directory at time of invocation
- The input directory that the program will search for whatever
input files are specified. If those files are not found in
this directory, the programs behavior is governed by the
value of the --interactive flag
* --fileList file1 [file2] [file3] ...
REQUIRED: NO
DEFAULT: The list of all files in --directory (non-recursive)
- The list of files that the program should attempt to
process for output. Most output functions (with the
exception of the function designed for MOPITT CO) are
designed to accept an arbitrary number of inputs and only
use that data which fits the requirements.
- filtering out files a priori which cannot contain the
information required, for instance files known to be on the
wrong date, saves computational power. It is advisable to
use this parameter to include only those files which may
contain the desired information wherever possible.
- Locations of data (current as of 01/12/12
MOPITT - <http://eosweb.larc.nasa.gov/HPDOCS/datapool/>
NASA OMI - <http://mirador.gsfc.nasa.gov/cgi-bin/mirador/collectionlist.pl?keyword=omno2>
KNMI OMI - <http://www.temis.nl/airpollution/no2col/data/omi/data_v2/>
* --filetype {HDFknmiomil2, HDFmopittl2, HDFnasaomil2}
REQUIRED: YES
DEFAULT: N/A
- The type of file we're attempting to read in. Must be one
of the options listed above. This must match the format of
the file (currently only the standard level 2 files for each
of the above listed instruments/retrievals are supported).
- This is probably the easiest part of the program to extend,
so if you're looking to use a particular filetype that isn't
supported have a look at parse_geo.py and look at possibly
extending it to include your filetype.
* --gridProj {latlon, lcc2par}
REQUIRED: YES
DEFAULT: N/A
- The grid projection used to define the target grid (the grid
that we wish to regrid our data to). Must be one of the above
options. Further details on these options are as follows:
latlon - The Plate Caree projection, also known as
the "unprojected" projection. x and y are
mapped directly to longitude and latitude,
respectively.
REQUIRED PARAMETERS:
xCell - The size of a gridcell in the x
(longitude) direction. In degrees.
yCell - The size of a gridcell in the y
(latitude) direction. In degrees.
xOrig - The longitude of the lower-left
corner of the domain
yOrig - The latitude of the lower-left
corner of the domain
nRows - The number of rows in the grid.
nCols - The number of columns in the grid.
lcc2par - The Lambert Conic Conformal projection (2
parallel construction). x and y are
transformed and scaled, then mapped to
latitude and longitude via the projection.
The projection parameters must be accurate to
get the correct output grid. All required
parameters are available in the GRIDDESC file
associated with the MM3 modeling system. A
description of the GRIDDESC format can be
found at
http://www.baronams.com/products/ioapi/GRIDDESC.html
REQUIRED PARAMETERS:
stdPar1 - One of the 2 standard parallels
used to define the Lambert Conic
Conformal projection. Must be a
valid latitude, in degrees.
stdPar2 - The second standard parallel used
to define the Lambert Conic
Conformal projection. Set this
equal to the same value as stdPar1
if the single-parallel form of the
projection is being used. Must be a
valid latitude in degrees.
refLat - The reference latitude upon which
the projection is centered. This is
the YCENT value in the GRIDDESC
file. In degrees
refLon - The reference longitude upon which
the projection is centered. This
is BOTH the XCENT and PROJ_GAMMA
values in the GRIDDESC file. If
these values are not identical, do
not use this function. In degrees.
xOrig - The location of the origin in
projected x coordinates. This
is the XORIG value in the GRIDDESC
file. In same units as earthRadius.
yOrig - The location of the origin in
projected y coordinates. This
is the YORIG value in the GRIDDESC
file. In same units as earthRadius.
xCell - The x dimension of a cell, in
projected coordinates. IN the same
units as earthRadius. This is the
XCELL value in the GRIDDESC file
yCell - The y dimension of a cell, in
projected coordinates. In the same
units as earthRadius. This is the
YCELL value in the GRIDDESC file
nRows - The number of rows in the grid.
nCols - The number of columns in the grid.
earthRadius - The assumed radius of the Earth
(assumed spherical). Must match
units used for xCell and yCell.
* --projAttrs name1:value1 name2:value2 ...
REQUIRED: YES
DEFAULT: N/A
- The attributes required for the chosen projection. Must
have all the required attributes for that projection. The
required parameters for each projection are listed under
that projection's name above.
- Case-sensitive. Attribute names must EXACTLY match those
laid out above.
* --mapFunc {point_in_cell, regional_intersect}
REQUIRED: YES
DEFAULT: N/A
- The mapping function that will be used to assign pixels to
cell(s). Certain datasets have restrictions on which
mapping functions are usable.
- Also responsible for computing the "geometric weight" of
pixels. That is, this function computes the weight unique
to a cell/pixel combination. At present, neither of the
functions provide this functionality.
point_in_cell - Maps pixels defined by a single
lat/lon (usually the cell center) to whatever
grid cell that point lies inside in projected
space. This assigns each pixel to a unique
grid cell. Grid cells are open on the top and
right sides and closed on the left and lower
sides (with directions defined according to
the projected coordinate system). This
function is supported by all currently
available input filetypes. NOTE: for
filetypes where regional_intersect is
available it is strongly recommended to use
regional_intersect over point_in_cell.
regional_intersect - Maps pixels (as defined by
pairs of geocoordinates that nominally
correspond to pixel corners) to ALL gridcells
intersected. No geometric weights are
calculated. This function is currently
supported by HDFnasaomil2 and HDFknmiomil2
filetypes only. Makes several assumptions:
- Polar discontinuities not encountered
- Projection is NOT global
- grid is rectilinear in projected space.
- Pixels are convex polygons.
* --outFunc {OMNO2e_netCDF_avg,OMNO2e_wght_avg,unweighted_filtered_MOPITT_avg_netCDF}
REQUIRED: YES
DEFAULT: N/A
- The function that computes the output and writes the output
file. Functions are given significant freedom, but all
current functions take some kind of average and write it to
an output file.
- These functions are frequently designed around a particular
instrument or input format. Efforts are made to make them
as general as possible, but specialized output functions are
only guaranteed (and really should only be used) for the
parser types for which they have been designed. Where this
is the case, it is noted below.
- In all cases where a fieldname must be given for a
parameter it is the short name (the name used to access the
field through the parser) that must be given. The
fieldnames must correspond to the official field names
in the data file. These can usually be found in the data
documentation. Documentation for a few commonly processed
formats can be found at:
OMI (KNMI) - <http://www.temis.nl/docs/OMI_NO2_HE5_1.0.2.pdf>
OMI (NASA) - <http://toms.gsfc.nasa.gov/omi/no2/OMNO2_data_product_specification.pdf>
MOPITT - <http://www.acd.ucar.edu/mopitt/v5_users_guide_beta.pdf>
OMNO2e_netCDF_avg - Averaging algorithm based on the
NASA OMI level 2 to level 3 processing
algorithm. Designed for the OMI level 2
filetypes (HDFnasaomil2 and HDFknmiomil2).
Use with other filetypes is of questionable
utility.
Outputs results to a netCDF file.
Further details available in the
official NASA documentation located at
<http://disc.sci.gsfc.nasa.gov/Aura/data-holdings/OMI/omno2e_v003.shtml>
Assumptions:
- Data is at most 3 dimensional.
- Invalid pixels are marked with an overall
quality flag.
- Timestamps are in the TAI93 format.
OMNO2e_wght_avg - Weighted averaging algorithm based
on the NASA algorithm used to process OMI from
level 2 to level 3. Designed for the OMI
level 2 filetypes (currently HDFnasaomil2 and
HDFknmiomil2). Use with other filetypes is of
questionable utility.
Outputs results to a CSV file. Does
not filter based on time as OMNO2e_netcdf_avg
does.
CSV file is designed such that if it is put
into an array from top left to bottom right,
the indices of the values will be correct.
Further details available in the official NASA
documentation located at
<http://disc.sci.gsfc.nasa.gov/Aura/data-holdings/OMI/omno2e_v003.shtml>
Assumptions:
- Data is 2 dimensional
- Invalid pixels are marked with an overall
quality flag
- Timestamps within the file should be in the
TAI-93 standard
unweighted_filtered_MOPITT_avg_netCDF - Averaging
algorithm based on the NASA algorithm for
processing level 2 MOPITT CO data to level 3
MOPITT CO data. Designed for the MOPITT level
2 filetypes (HDFmopittl2 only at present).
Use with other filetypes is discouraged.
Outputs resuts to a netCDF file.
IMPORTANT: Only 1 input file may be used with
this function. Conveniently, NASA currently
provides data in 1-day granules.
Further details available in the official NASA
documentation:
Deeter, Merritt N (2009). MOPITT (Measurements
of Pollution in the Troposphere) Validated
Version 4 Product Users Guide. Available from
<http://www.acd.ucar.edu/mopitt/products.shtml>
Assumptions/caveats:
- All fields are filtered based on the
number of valid layers present in the
specified column field. Including 2D
fields.
- Timestamps are in the TAI93 format.
* --outFuncAttrs name1:value1 name2:value2
REQUIRED: YES
DEFAULT: N/A
- The attributes required for the chosen output function.
Must have all required attributes for that output function.
The required parameters for each projection are listed
above.
- If one of the "value" elements contains whitespace, enclose
the entire name:value pair in double quotes. For example:
the:full_monty <- okay
"the:full monty" <- okay
the:full monty <- not okay
- In many cases, a comma delimited list is requested. Make
sure that elements of the list are not separated by spaces.
For example:
pythons:EricIdle,JohnCleese <- okay
pythons:EricIdle, JohnCleese <- not okay
"pythons:Eric Idle,John Cleese" <- okay
"pythons:Eric Idle, John Cleese" <- not okay
- Case-sensitive. Attribute names must EXACTLY match those
laid out below.
- Below are the required parameters for each existing output
function. { } contain usable/recommended parameters for
applicable filetypes. These are based on the current
versions of the data at the time of writing and should be
double-checked:
OMNO2e_netCDF_avg -
overallQualFlag - The name of the field
containing the overall quality flag
for the pixels. This flag should be
true (1) for invalid pixels and false
(0) for valid pixels.
{ OMI KNMI - TroposphericColumnFlag
OMI NASA - vcdQualityFlags }
cloudFrac - The name of the field containing
the cloud fractions.
{ OMI KNMI - CloudFraction
OMI NASA - CloudFraction }
solarZenithAngle - The name of the field
containing the solar zenith angles in
degrees.
{ OMI KNMI - SolarZenithAngle
OMI NASA - SolarZenithAngle }
time - The name of the field containing the
timestamps. Timestamps are assumed to
be in the TAI-93 format.
{ OMI KNMI - Time
OMI NASA - Time }
longitude - The name of the field containing
the longitudes at cell centers.
Longitudes should be in degrees east.
{ OMI KNMI - Longitude
OMI NASA - Longitude }
inFieldNames - The names of the fields desired
to be output. Input as comma
delimited list.
outFieldNames - The names of the output
variables (even if they are to be the
same as input variables). Should be a
comma-delimited list co-indexed to
inFieldNames
outUnits - The units of the variables to be
written out. Should be a
comma-delimited list co-indexed to
inFieldNames
extraDimLabel - Label for the extra dimension
(should the variable have an extra
dimension). Ignored in the case of a
2D variable. Should be a
comma-delimited list co-indexed to
inFieldNames
extraDimSizes - The size of the extra
dimensions (should the variable have
an extra dimension). For 2D
variables, must be set to 0. (zero)
Should be a comma-delimited list
co-indexed to inFieldNames.
timeComparison - Must be set to either "local"
or "UTC". Determines how the file
timestamps are compared to the
start/stop time. If set to "local",
then the file timestamps are converted
to local time on a pixel-by-pixel
basis (using longitude to estimate
time zone) before being compared to
time boundaries. If set to "UTC" the
file timestamps (which are assumed to
be in UTC) are compared against the
start/stop time directly.
timeStart - The earliest time for which data
should be recorded into the output
file. All times in input files before
this time will be filtered out. Must
be in the format:
hh:mm:ss_MM-DD-YYYY
timeStop - The latest time for which data
should be recorded into the output
files. All times in input files
after this time will be filtered out.
Must be in the format:
hh:mm:ss_MM-DD-YYYY
cloudFractUpperCutoff - The maximum cloud
fraction to allow before excluding
pixel from average. Suggested value
from NASA is 0.3
solarZenAngUpperCutoff - The maximum solar
zenith angle to allow before excluding
pixel from average. Suggested value
from NASA is 85. Must be in degrees.
pixIndXtrackAxis - The dimension order (0
based) of the "cross-track" dimension
(whichever dimension has size 60).
For all currently known cases set
equal to 1 (depends on the
construction of the parser function.
If you rewrite the parser, check
this).
fillVal - The value to use as a fill value in
the output netCDF file. This value
will replace any missing or invalid
output values.
OMNO2e_wght_avg -
toAvg - The name of the field to be averaged
overallQualFlag - The name of the field
containing the overall quality flag
for the pixels. Flag should be true (1)
for invalid pixels and false (0) for
valid pixels.
{ OMI KNMI - TroposphericColumnFlag
OMI NASA - vcdQualityFlags }
cloudFrac - The name of the field containing
the cloud fraction.
{ OMI KNMI - CloudFraction
OMI NASA - CloudFraction }
solarZenithAngle - The name of the field
containing the solar zenith angles.
{ OMI KNMI - SolarZenithAngle
OMI NASA - SolarZenithAngle }
cloudFractUpperCutoff - The maximum cloud
fraction to allow before excluding
pixel from average. Suggested value
from NASA is 0.3
solarZenAngUpperCutoff - The maximum solar
zenith angle to allow before excluding
pixel from average. Suggested value
from NASA is 85. Must be in degrees.
pixIndXtrackAxis - The dimension order (0
based) of the "cross-track" dimension
(whichever dimension has size 60).
For all currently known cases should
be 1 (may change in future versions of
OMI products).
fillVal - The value to use as a fill value in
the output netCDF file. This value
will replace any missing or invalid
output values.
unweighted_filtered_MOPITT_avg_netCDF -
time - The name of the field containing
timestamps. Timestamps are assumed to
be in the TAI-93 format.
{ MOPITT - Time }
longitude - The name of the field containing
the longitudes at cell centers.
Longitudes should be in degrees east.
{ MOPITT - Longitude }
inFieldNames - The names of the fields desired
to be output. Input as comma
delimited list.
outFieldNames - The names of the output
variables (even if they are to be the
same as input variables). Should be a
comma-delimited list co-indexed to
inFieldNames
outUnits - The units of the variables to be
written out. Should be a
comma-delimited list co-indexed to
inFieldNames
logNormal - List of boolean strings that
specify how to take the averages of
the corresponding fields. If the
string is "True" that field is
averaged assuming a lognormal
distribution. If the string is
"False" that field is averaged
assuming a normal distribution.
Official documentation (linked above)
has further information on when
log-average is appropriate. Should be
a comma-delimited list co-indexed to
inFieldNames
dimLabels - List of names of the extra
dimensions in the output file. Must
be a semicolon-delimited list of
comma-delimited lists of labels.
Fields with no extra dimensions may
be left blank. For example, if
there are four inFields, the first
and third of which have no extra
dimensions, the second of which has
one ("foo"), and the fourth has two
("foo" and "bar"), the dimLabels
entry should look like this:
;foo;;foo,bar
The outer (semicolon-delimited) list
must be co-indexed to inFieldNames.
dimSizes - List of the sizes of the extra
dimensions in the output file. Must
be a semicolon-delimited list of
comma-delimited lists of integers.
Fields with no extra dimensions may
be left blank. For example, if there
are four inFields, the first and
third of which have no extra
dimensions, the second of which has
one (which has length 4), and the
fourth has two (which have lengths
four and five, respectively), the
dimSizes entry should look like this:
;4;;4,5
The outer (semicolon-delimited list
must be co-indexed to inFieldNames
and each inner (comma-delimited) list
should be the same size as the
corresponding sublist in dimLabels.
timeStart - The earliest time for which data
should be recorded into the output
file. All times before this time in
the input file(s) will be filtered
out. Must be in the format:
hh:mm:ss_MM-DD-YYYY
timeStop - The latest time for which data
should be recorded into the output
file. All times after this time in
the input file(s) will be filtered
out. Must be in the format:
hh:mm:ss_MM-DD-YYYY
timeComparison - Must be set to either "local"
or "UTC". Determines how the file
timestamps are compared to the
start/stop time. If set to "local",
then the file timestamps are converted
to local time on a pixel-by-pixel
basis (using longitude to estimate
time zone) before being compared to
time boundaries. If set to "UTC" the
file timestamps (which are assumed to
be in UTC) are compared against the
start/stop time directly.
fillVal - The value to use as a fill value in
the output netCDF file. This value
will replace any missing or invalid
output values.
solZenAngCutoff - The solar zenith angle that
defines the day to night transition
(we use the SZA to separate day and
night pixels, which should not be
averaged together). The geometric
value here would be 90. Recommended
value is 85. In degrees.
solZenAng - The name of the field containing
the solar zenith angle (in degrees).
{ MOPITT - Solar Zenith Angle }
dayTime - Boolean variable that indicates
whether the output file should contain
values from day or night. If set to
"True" the output file will have
daylight values. If set to "False"
the output file will have night
values.
surfTypeField - The name of the field
containing the surface type index.
{ MOPITT - Surface Index }
colMeasField - The name of the field
containing the column measurement that
will be used to determine how many
valid layers are present in the cell.
This field must be 4 dimensional, with
the first extra dimension being the
level and the first element of the
second extra dimension containing
NaN's at the appropriate levels.
{ MOPITT - Retrieved CO Mixing Ratio Profile }
* --outDirectory /path/to/output/directory
REQUIRED: YES
DEFAULT: N/A
- The output directory to which the output file(s) should be
written. Make sure that you have write permissions to this
directory (the program will complain and quit out if you do
not).
* --outFileName FileName
REQUIRED: NO
DEFAULT: output1
- The name of the output file itself. User is responsible for
adding any file extensions here (IE .nc if it's a netCDF,
.txt if it's an ASCII). Output will be written in
outDirectory under this name.
* --includeGrid GridFileName
REQUIRED: NO
DEFAULT: N/A
- Supply this flag along with the name of a file (in the output directory)
to which to write out the latitudes and longitudes of the gridcells
defined by the selected projection.
* --verbose {True,False}
REQUIRED: NO
DEFAULT: True
- Determines how much command-line output the software
provides while running. The default behavior (--verbose
True) provide command line updates for most major
subprocesses inside the software. Setting "False" here will
cause the software to be completely silent while running.
* --interactive {True,False}
REQUIRED: NO
DEFAULT: False
- Determines how the program will handle invalid/nonexistent
files. Under the default behavior (--interactive False)
the software will automatically ignore any file it can't
process and continue processing any other files in
--fileList. If set to True, exectuion will be suspended and
the user will be given several options when an invalid file
is encountered.
* --AttributeHelp ProjectionName/OutputFunctionName [...]
REQUIRED: NO
DEFAULT: N/A
- Prints a message explaining the required attributes for a
given output function or projection and then exits the
program.
Wisconsin Horizontal Interpolation Program for Satellites (WHIPS)
=================================================================
**Project Title:** WHIPS
**Purpose of Project:** Provide a well-documented, easy-to-use general-purpose
processing module for processing satellite data
**Version:** 0.2.82 (2/3/2012)
**Authors:** oberman, maki, strom
**Contact:** oberman@wisc.edu
QUICK START
===========
1. Install the program and run the built-in test module to confirm
that it is working properly. Installation instructions can be found
in the file INSTALL.txt
2. Download whatever data you plan to process. Currently, the program
is designed to process OMI NO2 DOMINO level 2 data, OMI NO2 NASA level
2 data and MOPITT CO data. See the detailed documentation for the
--filelist argument for locations of data.
3. Navigate to the folder where process.py is located (or add it to
your path) and invoke it as:
process.py --help
4. Follow the on-screen instructions, adding each of the required
parameters. If you need help with the projection attributes or the
output function attributes, invoke the built-in help as:
process.py --AttributeHelp <function_name>
For detailed explanations of all parameters and attributes, see the
"Parameter Details" section below.
5. Invoke process.py once for each output file you'd like to create.
Note that the software creates output files with only a single
timestep, so you'll need to invoke the command once for each timestep
(IE if you want a month with timesteps every day, you'll probably want
to write a shell script that calls the command once for each day)
6. Additionally, you may create a grid file for your chosen grid by
including the --includeGrid flag followed by the desired filename.
A file containing the gridcells used by the projection will be written
to the same directory as the standard output file.
7. Concatenate your outputs if desired (the authors recommend the NCO
operators at http://nco.sourceforge.net/ if you're using a netCDF
output format) and carry on!
EXAMPLE/TEMPLATE INVOCATIONS
============================
For clarity and readability, line continuation characters are used to
place each attribute on a separate line. It is not required to break
up attributes like this, but the command line needs to see the
invocation as a single command, so if you want to break it onto
multple lines you must use line-continuation characters.
1. Process MOPITT level 2 CO data, (Version 5)
----------------------------------------------
- processes a single file
- Uses a 36km lambert conic conformal grid centered over North
America
- Writes out a 2D, 3D, and 4D parameter from the file
process.py \
--directory /path/to/input/files/ \
--filetype HDFmopittl2 \
--fileList MOP02T-20050106-L2V10.1.1.prov.hdf \
--gridProj lcc2par \
--projAttrs xOrig:-2916000 yCell:36000 refLon:-97 refLat:40 \
nCols:162 nRows:126 stdPar2:45 stdPar1:33 xCell:36000 \
earthRadius:6370000 yOrig:-2268000\
--mapFunc point_in_cell \
--outFunc unweighted_filtered_MOPITT_avg_netCDF \
--outFuncAttrs time:Time longitude:Longitude \
"inFieldNames:Time,Retrieved CO Mixing Ratio Profile,Retrieved CO Surface Mixing Ratio" \
outFieldNames:time,COprof,COsurf outUnits:TAI93,ppbv,ppbv \
logNormal:False,True,True "dimLabels:;layer,valOrStdDev;valOrStdDev" \
"dimSizes:;9,2;2" "timeStart:00:00:00_01-06-2005" \
"timeStop:23:59:59_01-06-2005" timeComparison:UTC \
fillVal:-9999.0 \
solZenAngCutoff:85 \
"solZenAng:Solar Zenith Angle" dayTime:True \
"surfTypeField:Surface Index" \
"colMeasField:Retrieved CO Mixing Ratio Profile" \
--outDirectory /path/to/output/directory/ \
--outFileName MOPITT_v5_20050106_daytime_CONUS36km_test.nc \
--verbose True \
--interactive True
2. Process OMI level 2 DOMINO NO2 data, (Version 5)
---------------------------------------------------
- processes any number of files
- Uses a 36km lambert conic conformal grid centered over North
America
- Writes out a 2D, and 3D parameter from the file
process.py \
--directory /where/you/keep/the/files/ \
--filetype HDFknmiomil2 \
--fileList OMI-Aura_L2-OMDOMINO_2006m0701t0023-o10423_v003-2010m1008t224420.he5 \
OMI-Aura_L2-OMDOMINO_2006m0701t0112-o10424_v003-20120m1008t224845.he5 \
--gridProj lcc2par \
--projAttrs xOrig:-2916000 yCell:36000 refLon:-97 refLat:40 \
nCols:162 nRows:126 stdPar2:45 stdPar1:33 xCell:36000 \
earthRadius:6370000 yOrig:-2268000 \
--mapFunc regional_intersect \
--outFunc OMNO2e_netCDF_avg \
--outFuncAttrs overallQualFlag:TropColumnFlag "cloudFrac:Cloud Fraction" \
solarZenithAngle:SolarZenithAngle time:Time longitude:Longitude \
inFieldNames:Time,AveragingKernel,TroposphericVerticalColumn \
outFieldNames:time,avKern,tropVCD \
"outUnits:TAI93,unitless x 1000,molec/cm^2x1^-15" extraDimLabel:none,Layers,none \
extraDimSize:0,34,0 timeStart:00:00:00_07-01-2006 timeStop:23:59:59_07-01-2006 \
timeComparison:UTC fillVal:-9999 cloudFractUpperCutoff:0.3 \
solarZenAngUpperCutoff:85 pixIndXtrackAxis:1 \
--outDirectory /where/you/want/output \
--outFileName OMI_DOMINO_20060701_test.nc \
--includeGrid OMI_DOMINO_GridFileName.nc \
--verbose True \
--interactive True
PARAMETER DETAILS
=================
Each input attribute is explained here. Note that each output
function has a separate set of required inputs and that the
--outFuncAttrs is therefore broken down by output function. Make sure
you're referencing the details for the attributes relevant to the
function you want to use.
* --help
REQUIRED: NO
DEFAULT: N/A
- Display the onscreen help message and exit the program.
* --directory /path/to/input/directory
REQUIRED: NO
DEFAULT: the current working directory at time of invocation
- The input directory that the program will search for whatever
input files are specified. If those files are not found in
this directory, the programs behavior is governed by the
value of the --interactive flag
* --fileList file1 [file2] [file3] ...
REQUIRED: NO
DEFAULT: The list of all files in --directory (non-recursive)
- The list of files that the program should attempt to
process for output. Most output functions (with the
exception of the function designed for MOPITT CO) are
designed to accept an arbitrary number of inputs and only
use that data which fits the requirements.
- filtering out files a priori which cannot contain the
information required, for instance files known to be on the
wrong date, saves computational power. It is advisable to
use this parameter to include only those files which may
contain the desired information wherever possible.
- Locations of data (current as of 01/12/12
MOPITT - <http://eosweb.larc.nasa.gov/HPDOCS/datapool/>
NASA OMI - <http://mirador.gsfc.nasa.gov/cgi-bin/mirador/collectionlist.pl?keyword=omno2>
KNMI OMI - <http://www.temis.nl/airpollution/no2col/data/omi/data_v2/>
* --filetype {HDFknmiomil2, HDFmopittl2, HDFnasaomil2}
REQUIRED: YES
DEFAULT: N/A
- The type of file we're attempting to read in. Must be one
of the options listed above. This must match the format of
the file (currently only the standard level 2 files for each
of the above listed instruments/retrievals are supported).
- This is probably the easiest part of the program to extend,
so if you're looking to use a particular filetype that isn't
supported have a look at parse_geo.py and look at possibly
extending it to include your filetype.
* --gridProj {latlon, lcc2par}
REQUIRED: YES
DEFAULT: N/A
- The grid projection used to define the target grid (the grid
that we wish to regrid our data to). Must be one of the above
options. Further details on these options are as follows:
latlon - The Plate Caree projection, also known as
the "unprojected" projection. x and y are
mapped directly to longitude and latitude,
respectively.
REQUIRED PARAMETERS:
xCell - The size of a gridcell in the x
(longitude) direction. In degrees.
yCell - The size of a gridcell in the y
(latitude) direction. In degrees.
xOrig - The longitude of the lower-left
corner of the domain
yOrig - The latitude of the lower-left
corner of the domain
nRows - The number of rows in the grid.
nCols - The number of columns in the grid.
lcc2par - The Lambert Conic Conformal projection (2
parallel construction). x and y are
transformed and scaled, then mapped to
latitude and longitude via the projection.
The projection parameters must be accurate to
get the correct output grid. All required
parameters are available in the GRIDDESC file
associated with the MM3 modeling system. A
description of the GRIDDESC format can be
found at
http://www.baronams.com/products/ioapi/GRIDDESC.html
REQUIRED PARAMETERS:
stdPar1 - One of the 2 standard parallels
used to define the Lambert Conic
Conformal projection. Must be a
valid latitude, in degrees.
stdPar2 - The second standard parallel used
to define the Lambert Conic
Conformal projection. Set this
equal to the same value as stdPar1
if the single-parallel form of the
projection is being used. Must be a
valid latitude in degrees.
refLat - The reference latitude upon which
the projection is centered. This is
the YCENT value in the GRIDDESC
file. In degrees
refLon - The reference longitude upon which
the projection is centered. This
is BOTH the XCENT and PROJ_GAMMA
values in the GRIDDESC file. If
these values are not identical, do
not use this function. In degrees.
xOrig - The location of the origin in
projected x coordinates. This
is the XORIG value in the GRIDDESC
file. In same units as earthRadius.
yOrig - The location of the origin in
projected y coordinates. This
is the YORIG value in the GRIDDESC
file. In same units as earthRadius.
xCell - The x dimension of a cell, in
projected coordinates. IN the same
units as earthRadius. This is the
XCELL value in the GRIDDESC file
yCell - The y dimension of a cell, in
projected coordinates. In the same
units as earthRadius. This is the
YCELL value in the GRIDDESC file
nRows - The number of rows in the grid.
nCols - The number of columns in the grid.
earthRadius - The assumed radius of the Earth
(assumed spherical). Must match
units used for xCell and yCell.
* --projAttrs name1:value1 name2:value2 ...
REQUIRED: YES
DEFAULT: N/A
- The attributes required for the chosen projection. Must
have all the required attributes for that projection. The
required parameters for each projection are listed under
that projection's name above.
- Case-sensitive. Attribute names must EXACTLY match those
laid out above.
* --mapFunc {point_in_cell, regional_intersect}
REQUIRED: YES
DEFAULT: N/A
- The mapping function that will be used to assign pixels to
cell(s). Certain datasets have restrictions on which
mapping functions are usable.
- Also responsible for computing the "geometric weight" of
pixels. That is, this function computes the weight unique
to a cell/pixel combination. At present, neither of the
functions provide this functionality.
point_in_cell - Maps pixels defined by a single
lat/lon (usually the cell center) to whatever
grid cell that point lies inside in projected
space. This assigns each pixel to a unique
grid cell. Grid cells are open on the top and
right sides and closed on the left and lower
sides (with directions defined according to
the projected coordinate system). This
function is supported by all currently
available input filetypes. NOTE: for
filetypes where regional_intersect is
available it is strongly recommended to use
regional_intersect over point_in_cell.
regional_intersect - Maps pixels (as defined by
pairs of geocoordinates that nominally
correspond to pixel corners) to ALL gridcells
intersected. No geometric weights are
calculated. This function is currently
supported by HDFnasaomil2 and HDFknmiomil2
filetypes only. Makes several assumptions:
- Polar discontinuities not encountered
- Projection is NOT global
- grid is rectilinear in projected space.
- Pixels are convex polygons.
* --outFunc {OMNO2e_netCDF_avg,OMNO2e_wght_avg,unweighted_filtered_MOPITT_avg_netCDF}
REQUIRED: YES
DEFAULT: N/A
- The function that computes the output and writes the output
file. Functions are given significant freedom, but all
current functions take some kind of average and write it to
an output file.
- These functions are frequently designed around a particular
instrument or input format. Efforts are made to make them
as general as possible, but specialized output functions are
only guaranteed (and really should only be used) for the
parser types for which they have been designed. Where this
is the case, it is noted below.
- In all cases where a fieldname must be given for a
parameter it is the short name (the name used to access the
field through the parser) that must be given. The
fieldnames must correspond to the official field names
in the data file. These can usually be found in the data
documentation. Documentation for a few commonly processed
formats can be found at:
OMI (KNMI) - <http://www.temis.nl/docs/OMI_NO2_HE5_1.0.2.pdf>
OMI (NASA) - <http://toms.gsfc.nasa.gov/omi/no2/OMNO2_data_product_specification.pdf>
MOPITT - <http://www.acd.ucar.edu/mopitt/v5_users_guide_beta.pdf>
OMNO2e_netCDF_avg - Averaging algorithm based on the
NASA OMI level 2 to level 3 processing
algorithm. Designed for the OMI level 2
filetypes (HDFnasaomil2 and HDFknmiomil2).
Use with other filetypes is of questionable
utility.
Outputs results to a netCDF file.
Further details available in the
official NASA documentation located at
<http://disc.sci.gsfc.nasa.gov/Aura/data-holdings/OMI/omno2e_v003.shtml>
Assumptions:
- Data is at most 3 dimensional.
- Invalid pixels are marked with an overall
quality flag.
- Timestamps are in the TAI93 format.
OMNO2e_wght_avg - Weighted averaging algorithm based
on the NASA algorithm used to process OMI from
level 2 to level 3. Designed for the OMI
level 2 filetypes (currently HDFnasaomil2 and
HDFknmiomil2). Use with other filetypes is of
questionable utility.
Outputs results to a CSV file. Does
not filter based on time as OMNO2e_netcdf_avg
does.
CSV file is designed such that if it is put
into an array from top left to bottom right,
the indices of the values will be correct.
Further details available in the official NASA
documentation located at
<http://disc.sci.gsfc.nasa.gov/Aura/data-holdings/OMI/omno2e_v003.shtml>
Assumptions:
- Data is 2 dimensional
- Invalid pixels are marked with an overall
quality flag
- Timestamps within the file should be in the
TAI-93 standard
unweighted_filtered_MOPITT_avg_netCDF - Averaging
algorithm based on the NASA algorithm for
processing level 2 MOPITT CO data to level 3
MOPITT CO data. Designed for the MOPITT level
2 filetypes (HDFmopittl2 only at present).
Use with other filetypes is discouraged.
Outputs resuts to a netCDF file.
IMPORTANT: Only 1 input file may be used with
this function. Conveniently, NASA currently
provides data in 1-day granules.
Further details available in the official NASA
documentation:
Deeter, Merritt N (2009). MOPITT (Measurements
of Pollution in the Troposphere) Validated
Version 4 Product Users Guide. Available from
<http://www.acd.ucar.edu/mopitt/products.shtml>
Assumptions/caveats:
- All fields are filtered based on the
number of valid layers present in the
specified column field. Including 2D
fields.
- Timestamps are in the TAI93 format.
* --outFuncAttrs name1:value1 name2:value2
REQUIRED: YES
DEFAULT: N/A
- The attributes required for the chosen output function.
Must have all required attributes for that output function.
The required parameters for each projection are listed
above.
- If one of the "value" elements contains whitespace, enclose
the entire name:value pair in double quotes. For example:
the:full_monty <- okay
"the:full monty" <- okay
the:full monty <- not okay
- In many cases, a comma delimited list is requested. Make
sure that elements of the list are not separated by spaces.
For example:
pythons:EricIdle,JohnCleese <- okay
pythons:EricIdle, JohnCleese <- not okay
"pythons:Eric Idle,John Cleese" <- okay
"pythons:Eric Idle, John Cleese" <- not okay
- Case-sensitive. Attribute names must EXACTLY match those
laid out below.
- Below are the required parameters for each existing output
function. { } contain usable/recommended parameters for
applicable filetypes. These are based on the current
versions of the data at the time of writing and should be
double-checked:
OMNO2e_netCDF_avg -
overallQualFlag - The name of the field
containing the overall quality flag
for the pixels. This flag should be
true (1) for invalid pixels and false
(0) for valid pixels.
{ OMI KNMI - TroposphericColumnFlag
OMI NASA - vcdQualityFlags }
cloudFrac - The name of the field containing
the cloud fractions.
{ OMI KNMI - CloudFraction
OMI NASA - CloudFraction }
solarZenithAngle - The name of the field
containing the solar zenith angles in
degrees.
{ OMI KNMI - SolarZenithAngle
OMI NASA - SolarZenithAngle }
time - The name of the field containing the
timestamps. Timestamps are assumed to
be in the TAI-93 format.
{ OMI KNMI - Time
OMI NASA - Time }
longitude - The name of the field containing
the longitudes at cell centers.
Longitudes should be in degrees east.
{ OMI KNMI - Longitude
OMI NASA - Longitude }
inFieldNames - The names of the fields desired
to be output. Input as comma
delimited list.
outFieldNames - The names of the output
variables (even if they are to be the
same as input variables). Should be a
comma-delimited list co-indexed to
inFieldNames
outUnits - The units of the variables to be
written out. Should be a
comma-delimited list co-indexed to
inFieldNames
extraDimLabel - Label for the extra dimension
(should the variable have an extra
dimension). Ignored in the case of a
2D variable. Should be a
comma-delimited list co-indexed to
inFieldNames
extraDimSizes - The size of the extra
dimensions (should the variable have
an extra dimension). For 2D
variables, must be set to 0. (zero)
Should be a comma-delimited list
co-indexed to inFieldNames.
timeComparison - Must be set to either "local"
or "UTC". Determines how the file
timestamps are compared to the
start/stop time. If set to "local",
then the file timestamps are converted
to local time on a pixel-by-pixel
basis (using longitude to estimate
time zone) before being compared to
time boundaries. If set to "UTC" the
file timestamps (which are assumed to
be in UTC) are compared against the
start/stop time directly.
timeStart - The earliest time for which data
should be recorded into the output
file. All times in input files before
this time will be filtered out. Must
be in the format:
hh:mm:ss_MM-DD-YYYY
timeStop - The latest time for which data
should be recorded into the output
files. All times in input files
after this time will be filtered out.
Must be in the format:
hh:mm:ss_MM-DD-YYYY
cloudFractUpperCutoff - The maximum cloud
fraction to allow before excluding
pixel from average. Suggested value
from NASA is 0.3
solarZenAngUpperCutoff - The maximum solar
zenith angle to allow before excluding
pixel from average. Suggested value
from NASA is 85. Must be in degrees.
pixIndXtrackAxis - The dimension order (0
based) of the "cross-track" dimension
(whichever dimension has size 60).
For all currently known cases set
equal to 1 (depends on the
construction of the parser function.
If you rewrite the parser, check
this).
fillVal - The value to use as a fill value in
the output netCDF file. This value
will replace any missing or invalid
output values.
OMNO2e_wght_avg -
toAvg - The name of the field to be averaged
overallQualFlag - The name of the field
containing the overall quality flag
for the pixels. Flag should be true (1)
for invalid pixels and false (0) for
valid pixels.
{ OMI KNMI - TroposphericColumnFlag
OMI NASA - vcdQualityFlags }
cloudFrac - The name of the field containing
the cloud fraction.
{ OMI KNMI - CloudFraction
OMI NASA - CloudFraction }
solarZenithAngle - The name of the field
containing the solar zenith angles.
{ OMI KNMI - SolarZenithAngle
OMI NASA - SolarZenithAngle }
cloudFractUpperCutoff - The maximum cloud
fraction to allow before excluding
pixel from average. Suggested value
from NASA is 0.3
solarZenAngUpperCutoff - The maximum solar
zenith angle to allow before excluding
pixel from average. Suggested value
from NASA is 85. Must be in degrees.
pixIndXtrackAxis - The dimension order (0
based) of the "cross-track" dimension
(whichever dimension has size 60).
For all currently known cases should
be 1 (may change in future versions of
OMI products).
fillVal - The value to use as a fill value in
the output netCDF file. This value
will replace any missing or invalid
output values.
unweighted_filtered_MOPITT_avg_netCDF -
time - The name of the field containing
timestamps. Timestamps are assumed to
be in the TAI-93 format.
{ MOPITT - Time }
longitude - The name of the field containing
the longitudes at cell centers.
Longitudes should be in degrees east.
{ MOPITT - Longitude }
inFieldNames - The names of the fields desired
to be output. Input as comma
delimited list.
outFieldNames - The names of the output
variables (even if they are to be the
same as input variables). Should be a
comma-delimited list co-indexed to
inFieldNames
outUnits - The units of the variables to be
written out. Should be a
comma-delimited list co-indexed to
inFieldNames
logNormal - List of boolean strings that
specify how to take the averages of
the corresponding fields. If the
string is "True" that field is
averaged assuming a lognormal
distribution. If the string is
"False" that field is averaged
assuming a normal distribution.
Official documentation (linked above)
has further information on when
log-average is appropriate. Should be
a comma-delimited list co-indexed to
inFieldNames
dimLabels - List of names of the extra
dimensions in the output file. Must
be a semicolon-delimited list of
comma-delimited lists of labels.
Fields with no extra dimensions may
be left blank. For example, if
there are four inFields, the first
and third of which have no extra
dimensions, the second of which has
one ("foo"), and the fourth has two
("foo" and "bar"), the dimLabels
entry should look like this:
;foo;;foo,bar
The outer (semicolon-delimited) list
must be co-indexed to inFieldNames.
dimSizes - List of the sizes of the extra
dimensions in the output file. Must
be a semicolon-delimited list of
comma-delimited lists of integers.
Fields with no extra dimensions may
be left blank. For example, if there
are four inFields, the first and
third of which have no extra
dimensions, the second of which has
one (which has length 4), and the
fourth has two (which have lengths
four and five, respectively), the
dimSizes entry should look like this:
;4;;4,5
The outer (semicolon-delimited list
must be co-indexed to inFieldNames
and each inner (comma-delimited) list
should be the same size as the
corresponding sublist in dimLabels.
timeStart - The earliest time for which data
should be recorded into the output
file. All times before this time in
the input file(s) will be filtered
out. Must be in the format:
hh:mm:ss_MM-DD-YYYY
timeStop - The latest time for which data
should be recorded into the output
file. All times after this time in
the input file(s) will be filtered
out. Must be in the format:
hh:mm:ss_MM-DD-YYYY
timeComparison - Must be set to either "local"
or "UTC". Determines how the file
timestamps are compared to the
start/stop time. If set to "local",
then the file timestamps are converted
to local time on a pixel-by-pixel
basis (using longitude to estimate
time zone) before being compared to
time boundaries. If set to "UTC" the
file timestamps (which are assumed to
be in UTC) are compared against the
start/stop time directly.
fillVal - The value to use as a fill value in
the output netCDF file. This value
will replace any missing or invalid
output values.
solZenAngCutoff - The solar zenith angle that
defines the day to night transition
(we use the SZA to separate day and
night pixels, which should not be
averaged together). The geometric
value here would be 90. Recommended
value is 85. In degrees.
solZenAng - The name of the field containing
the solar zenith angle (in degrees).
{ MOPITT - Solar Zenith Angle }
dayTime - Boolean variable that indicates
whether the output file should contain
values from day or night. If set to
"True" the output file will have
daylight values. If set to "False"
the output file will have night
values.
surfTypeField - The name of the field
containing the surface type index.
{ MOPITT - Surface Index }
colMeasField - The name of the field
containing the column measurement that
will be used to determine how many
valid layers are present in the cell.
This field must be 4 dimensional, with
the first extra dimension being the
level and the first element of the
second extra dimension containing
NaN's at the appropriate levels.
{ MOPITT - Retrieved CO Mixing Ratio Profile }
* --outDirectory /path/to/output/directory
REQUIRED: YES
DEFAULT: N/A
- The output directory to which the output file(s) should be
written. Make sure that you have write permissions to this
directory (the program will complain and quit out if you do
not).
* --outFileName FileName
REQUIRED: NO
DEFAULT: output1
- The name of the output file itself. User is responsible for
adding any file extensions here (IE .nc if it's a netCDF,
.txt if it's an ASCII). Output will be written in
outDirectory under this name.
* --includeGrid GridFileName
REQUIRED: NO
DEFAULT: N/A
- Supply this flag along with the name of a file (in the output directory)
to which to write out the latitudes and longitudes of the gridcells
defined by the selected projection.
* --verbose {True,False}
REQUIRED: NO
DEFAULT: True
- Determines how much command-line output the software
provides while running. The default behavior (--verbose
True) provide command line updates for most major
subprocesses inside the software. Setting "False" here will
cause the software to be completely silent while running.
* --interactive {True,False}
REQUIRED: NO
DEFAULT: False
- Determines how the program will handle invalid/nonexistent
files. Under the default behavior (--interactive False)
the software will automatically ignore any file it can't
process and continue processing any other files in
--fileList. If set to True, exectuion will be suspended and
the user will be given several options when an invalid file
is encountered.
* --AttributeHelp ProjectionName/OutputFunctionName [...]
REQUIRED: NO
DEFAULT: N/A
- Prints a message explaining the required attributes for a
given output function or projection and then exits the
program.
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
WHIPS-1.0.0.tar.gz
(86.2 kB
view hashes)
