DESCRIPTION
Compare grib messages contained in two files. If some differences are found it fails returning an error code. Floating point values are compared exactly by default, different tolerance can be defined see -P -A -R. Default behaviour: absolute error=0, bit-by-bit compare, same order in files.
USAGE
grib_compare
...
OPTIONS
...
Compare files in which the messages are not in the same order. This option is time expensive.
...
All the keys in this list are skipped in the comparison. Bit-by-bit compare on.
...
Edition independent compare. It is used to compare grib edition 1 and 2.
...
Only the listed keys or namespaces (:n) are compared. The optional letter after the colon is used to force the type in the comparison: l->integer, d->float, s->string, n->namespace. See -a option. Incompatible with -H option.
...
First field to be processed.
...
Last field to be processed.
...
-c option modifier. The keys listed with the option -c will be added to the list of keys compared without -c.
...
Compare only message headers. Bit-by-bit compare on. Incompatible with -c option.
...
Compare floating point values using the relative error as tolerance. key1=relative_error will compare key1 using relative_error1. all=relative_error will compare all the floating point keys using relative_error. Default all=0.
...
Compare floating point values using the absolute error as tolerance. Default is absolute error=0
...
Compare data values using the packing error as tolerance.
...
Compare data values using factor multiplied by the tolerance specified in options -P -R -A.
...
Where clause. Messages are processed only if they match all the key/value constraints. A valid constraint is of type key=value or key!=value. For each key a string (key:s) or a double (key:d) or a long (key:l) type can be specified. Default type is string.
...
Force. Force the execution not to fail on error.
...
Version.
...
Does not fail when the message has wrong length
...
Verbose.
grib_compare examples
The default behaviour for grib_compare without any option is to perform a bit by bit comparison of the two messages. If the messages are found to be bitwise different then grib_compare switches to a "key based" mode to find out which coded keys are different. To see how grib_compare works we first set the shortName=2d (2 metre dew point temperature) in the file regular_latlon_surface.grib1
Then we can compare the two fields with grib_compare.Code Block > grib_set -s shortName=2d regular_latlon_surface.grib1 2d.grib1
In the output we see that the only "coded" key with different values in the two messages is indicatorOfParameter which is the relevant key for the parameter information. The comparison can be forced to be successful listing the keys with different values in the -b option.Code Block > grib_compare regular_latlon_surface.grib1 2d.grib1 -- GRIB #1 -- shortName=2t paramId=167 stepRange=0 levelType=sfc level=0 packingType=grid_simple gridType=regular_ll -- long [indicatorOfParameter]: [167] != [168]
Two grib messages can be very different because they have different edition, but they can contain the same identical information in the header and the same data. To see how grib_compare can help in comparing messages with different edition we doCode Block > grib_compare -b indicatorOfParameter regular_latlon_surface.grib1 2d.grib1
Then we compare the two fields with grib_compare.Code Block > grib_set edition=2 reduced_gaussian_model_level.grib1 reduced_gaussian_model_level.grib2
It is clear that the two messages are coded in a very different way. If we now add the -e option, the tool will compare only the higher level information common between the two messages.Code Block > grib_compare reduced_gaussian_model_level.grib1 reduced_gaussian_model_level.grib2 -- GRIB #1 -- shortName=t paramId=130 stepRange=0 levelType=ml level=1 packingType=grid_simple gridType=reduced_gg -- long [totalLength]: [10908] != [10996] long [editionNumber]: [1] != [2] long [section1Length]: [52] != [21] [table2Version] not found in 2nd field [gridDefinition] not found in 2nd field [indicatorOfParameter] not found in 2nd field [indicatorOfTypeOfLevel] not found in 2nd field [yearOfCentury] not found in 2nd field [unitOfTimeRange] not found in 2nd field [P1] not found in 2nd field [P2] not found in 2nd field [numberIncludedInAverage] not found in 2nd field [numberMissingFromAveragesOrAccumulations] not found in 2nd field [centuryOfReferenceTimeOfData] not found in 2nd field [reservedNeedNotBePresent] not found in 2nd field [perturbationNumber] not found in 2nd field [numberOfForecastsInEnsemble] not found in 2nd field [padding_local1_1] not found in 2nd field long [section2Length]: [896] != [17] [pvlLocation] not found in 2nd field [dataRepresentationType] not found in 2nd field long [latitudeOfFirstGridPoint]: [87864] != [87863799] long [latitudeOfLastGridPoint]: [-87864] != [-87863799] long [longitudeOfLastGridPoint]: [357188] != [357187500] [padding_grid4_1] not found in 2nd field long [section4Length]: [9948] != [770] [dataFlag] not found in 2nd field
The comparison is successful because the two messages contain the same information coded in two different ways. We can display the list of keys used by grib_compare adding the option -v (verbose).Code Block > grib_compare -e reduced_gaussian_model_level.grib1 reduced_gaussian_model_level.grib2 -- GRIB #1 -- shortName=t paramId=130 stepRange=0 levelType=ml level=1 packingType=grid_simple gridType=reduced_gg -- string [param]: [130.128] != [130]
For each key the type used in the comparison is reported and for the floating point keys also the tolerance used is printed.Code Block > grib_compare -ve reduced_gaussian_model_level.grib1 reduced_gaussian_model_level.grib2 reduced_gaussian_model_level.grib2 comparing centre as string comparing paramId as long comparing units as string comparing name as string comparing shortName as string comparing typeOfLevel as string comparing level as long comparing pv as double (184 values) tolerance=0 comparing bitmapPresent as long comparing latitudeOfFirstGridPointInDegrees as double (1 values) tolerance=0.0005 comparing longitudeOfFirstGridPointInDegrees as double (1 values) tolerance=0.0005 comparing latitudeOfLastGridPointInDegrees as double (1 values) tolerance=0.0005 comparing longitudeOfLastGridPointInDegrees as double (1 values) tolerance=0.0005 comparing iDirectionIncrementInDegrees is set to missing in both fields comparing N as long comparing iScansNegatively as long comparing jScansPositively as long comparing jPointsAreConsecutive as long comparing pl as long comparing gridType as string comparing packedValues as double (6114 values) tolerance=0 comparing domain as string comparing levtype as string comparing levelist as long comparing date as long comparing time as long comparing step as long comparing param as string -- GRIB #1 -- shortName=t paramId=130 stepRange=0 levelType=ml level=1 packingType=grid_simple gridType=reduced_gg -- string [param]: [130.128] != [130] comparing class as string comparing type as string comparing stream as string comparing expver as string
Some options are provided to compare only a set of keys in the messages. The option -H is used to compare only the headers coded in the message, it doesn't compare the data values. The option "-c key1:[l/d/s/n],key2:[l/d/s/n],... " can be used to compare a set of keys or namespaces. The letter after the colon is optional and it is used to force the type used in the comparison which is otherwise assumed to be the native type of the key. The possible types are:
- :l -> integer (C type long)
- :d -> floating point (C type double)
- :s -> string
- :n -> namespace.
We first compare the two files using the -H option (only headers are compared).Code Block set bitsPerValue=10; set values={1,2.5,3,4,5,6,70}; write "first.grib1"; set values={1,2.5,5,4,5,6,70}; write "second.grib1";
The comparison is successful because the data are not compared. To compare only the data we have to compare the "data namespace".Code Block > grib_compare -H first.grib1 second.grib1
The comparison is showing that one of seven values is different in a comparison with the (default) absolute tolerance=0. We can change the tolerance with the -A option:Code Block > grib_compare -c data:n first.grib1 second.grib1 -- GRIB #1 -- shortName=t paramId=130 stepRange=0 levelType=ml level=1 packingType=grid_simple gridType=reduced_gg -- double [packedValues]: 1 out of 7 different max absolute diff. = 2.0000000000000000e+00, relative diff. = 0.4 max diff. element 2: 3.00000000000000000000e+00 5.00000000000000000000e+00 tolerance=0.0000000000000000e+00 packingError: [0.0625005] [0.0625005] values max= [70] [70] min= [1] [1]
and we see that the comparison is successful if the absolute tolerance is set to 2. We can also set the relative tolerance for each key with the option -R:Code Block > grib_compare -A 2 -c data:n first.grib1 second.grib1
and we get again a successful comparison because the relative tolerance is bigger than the relative absolute difference of two corresponding values. Another possible choice for the tolerance is to be equal to the packingError, which is the error due to the packing algorithm. If we change the decimalPrecision of a packed field we introduce a packing error sometimes bigger than the original packing error.Code Block > grib_compare -R packedValues=0.4 -c data:n first.grib1 second.grib1
and we compare the two fields using the -P option (tolerance=packingError).Code Block > grib_set -s changeDecimalPrecision=0 first.grib1 third.grib1
the comparison is successful because their difference is within the biggest of the two packing error. With the option -P the comparison is failing only if the original data coded are different, not if the packing precision is changed. If we try again to compare the fields without the -P option:Code Block > grib_compare -P -c data:n first.grib1 third.grib1
we see that some values are different and that the maximum absolute differenc is close to the biggest packing error (max diff=0.48 packingError=0.5). The packing error was chosen to be 0.5 by setting decimalPrecision to 0 which means that we don't need to preserve any decimal figure.Code Block > grib_compare -c data:n first.grib1 third.grib1 -- GRIB #1 -- shortName=t paramId=130 stepRange=0 levelType=ml level=1 packingType=grid_simple gridType=reduced_gg -- double [packedValues]: 1 out of 7 different max absolute diff. = 5.0000000000000000e-01, relative diff. = 0.166667 max diff. element 1: 2.50000000000000000000e+00 3.00000000000000000000e+00 tolerance=0.0000000000000000e+00 packingError: [0.0625005] [0.5] values max= [70] [70] min= [1] [1]
When we already know that the fields are not numerically identical, but have similar statistical characteristics we can compare their statistics namespaces:
and we see that maximum, minimum, average, standard deviation, skewness and kurtosis are compared. While the values are different by 0.48 the statistics comparison shows that the difference in the statistical values is never bigger than 0.052Code Block > grib_compare -c statistics:n first.grib1 third.grib1 -- GRIB #1 -- shortName=t paramId=130 stepRange=0 levelType=ml level=1 packingType=grid_simple gridType=reduced_gg -- double [avg]: [1.30714285714285711748e+01] != [1.31428571428571423496e+01] absolute diff. = 0.0714286, relative diff. = 0.00543478 tolerance=0 double [sd]: [2.32907531796090587761e+01] != [2.32589679873534969090e+01] absolute diff. = 0.0317852, relative diff. = 0.00136471 tolerance=0 double [skew]: [2.02295027950165895447e+00] != [2.02385673400705590197e+00] absolute diff. = 0.000906455, relative diff. = 0.000447885 tolerance=0 double [kurt]: [2.12697527593972246507e+00] != [2.12906658242618895827e+00] absolute diff. = 0.00209131, relative diff. = 0.000982264 tolerance=0
The statistics namespace is available also for spherical harmonics data and provides information about the field in the geographic space computing them in the spectral space for performance reasons.Code Block > grib_compare -A 0.052 -c statistics:n first.grib1 third.grib1 -- GRIB #1 -- shortName=t paramId=130 stepRange=0 levelType=ml level=1 packingType=grid_simple gridType=reduced_gg -- double [avg]: [1.30714285714285711748e+01] != [1.31428571428571423496e+01] absolute diff. = 0.0714286, relative diff. = 0.00543478 tolerance=0.052
When a file contains several fields and some keys are different, it is useful to have a summary report of the keys found different in the messages. This can be obtained with the option -f. We change few keys in a file:
and comparing with the -f option:Code Block > grib_set -w typeOfLevel=surface -s step=48 tigge_pf_ecmwf.grib2 out.grib2
we get a list of all the different messages in the files and a summary report of the different keys.Code Block > grib_compare -f tigge_pf_ecmwf.grib2 out.grib2 -- GRIB #9 -- shortName=skt paramId=235 stepRange=96 levelType=sfc level=0 packingType=grid_simple gridType=regular_ll -- long [forecastTime]: [96] != [48] -- GRIB #10 -- shortName=sd paramId=228141 stepRange=96 levelType=sfc level=0 packingType=grid_simple gridType=regular_ll -- long [forecastTime]: [96] != [48] -- GRIB #11 -- shortName=sf paramId=228144 stepRange=0-96 levelType=sfc level=0 packingType=grid_simple gridType=regular_ll -- long [dayOfEndOfOverallTimeInterval]: [26] != [24] long [lengthOfTimeRange]: [96] != [48] ... output deleted ## ERRORS SUMMARY ####### ## ## Summary of different key values ## forecastTime ( 3 different ) ## dayOfEndOfOverallTimeInterval ( 11 different ) ## lengthOfTimeRange ( 11 different ) ## ## 14 different messages out of 38
We can change the order of the messages in a file using grib_copy with the -B option:
If we now compare the two files:Code Block > grib_copy -B typeOfLevel tigge_pf_ecmwf.grib2 out.grib2
the comparison is failing because of the different order of the messages. We can use the -r option to compare the files assuming that the messages are not in the same order:Code Block > grib_compare -f tigge_pf_ecmwf.grib2 out.grib2 -- GRIB #1 -- shortName=10u paramId=165 stepRange=96 levelType=sfc level=10 packingType=grid_simple gridType=regular_ll -- long [discipline]: [0] != [2] long [totalLength]: [1555] != [990] long [parameterCategory]: [2] != [0] long [parameterNumber]: [2] != [22] long [scaledValueOfFirstFixedSurface]: [10] != [0] long [typeOfSecondFixedSurface]: [255] != [106] scaleFactorOfSecondFixedSurface is set to missing in 1st field is not missing in 2nd field scaledValueOfSecondFixedSurface is set to missing in 1st field is not missing in 2nd field long [numberOfValues]: [684] != [239] double [referenceValue]: [-1.57229328155517578125e+01] != [4.15843811035156250000e+01] absolute diff. = 57.3073, relative diff. = 1.3781 tolerance=3.8147e-06 long [binaryScaleFactor]: [-10] != [-15] long [bitsPerValue]: [16] != [24] long [section6Length]: [6] != [92] long [bitMapIndicator]: [255] != [0] long [section7Length]: [1373] != [722] Different size for "codedValues" [684] [239] ... very long output
and we have a successful comparison because for each message in the first file an identical message is found in the second file. This option should be used carefully as it is very time expensive.Code Block > grib_compare -r tigge_pf_ecmwf.grib2 out.grib2