R help - Nov 2011 - Removal/selecting specific rows in a dataframe conditional on 2 columns

Dear list, 

After reading different mails, blogs, and tried a few different codes without
any success, I am asking your help!
I have the following data frame where each row represent a survey unit with the
following variables:
> names(RV09)
 [1] "record.t"  "trip"      "set"      
"month"     "stratum"   "NAFO"
 [7] "unit.area" "time"      "dur.set"  
"distance"  "operation" "mean.d"
[13] "min.d"     "max.d"     "temp.d"   
"slat"      "slong"     "spp"
[19] "number"    "weight"    "elat"     
"elong"

Each survey unit generates one set record, denoted by a 5 in column
"record.t". Each species identified in this particular survey unit
generates an additional set record, denoted by a 6.
> unique(RV09$record.t)
[1] 5 6

Each survey unit are identified by a specific "trip" and
"set" number, so if there is a 5 record type with no associated 6
records, it means that no species were observed in that survey unit. I would
like to be able to select all and only these survey units, which represent my
zeros.

So as an exemple, in this trip number 913, set 1, 3, and 4 would be part of my
"zeros" data.frame as they appear with no record.t 6, such that no
species were observed in this survey unit.
> head(RV09)
   record.t trip set month stratum NAFO unit.area time dur.set distance
585        5  913   1    10     351   3O       R31 1044      17        9
586        5  913   2    10     351   3O       R31 1440      17        9
587        6  913   2    10     351   3O       R31 1440      17        9
588        5  913   3    10     340   3O       Q31 1800      18        9
589        5  913   4    10     340   3O       Q32 2142      17        9

Any tips on how extract this "zero" data.frame in R? 
Thank you very much in advance!

Best,
~Aurelie


Aurelie Cosandey-Godin
Ph.D. student, Department of Biology
Industrial Graduate Fellow, WWF-Canada
Dalhousie University | Email: godina@dal.ca


	[[alternative HTML version deleted]]

Perhaps use tapply() to split by the survey unit and write a little
identity function that returns only those rows you want, then patch
them all back together with something like simplify2array().

Michael

On Tue, Nov 1, 2011 at 1:16 PM, Aurelie Cosandey Godin <godina at dal.ca>
wrote:
> Dear list,
>
> After reading different mails, blogs, and tried a few different codes
without any success, I am asking your help!
> I have the following data frame where each row represent a survey unit with
the following variables:
>
>> names(RV09)
> ?[1] "record.t" ?"trip" ? ? ?"set" ? ? ?
"month" ? ? "stratum" ? "NAFO"
> ?[7] "unit.area" "time" ? ? ?"dur.set" ?
"distance" ?"operation" "mean.d"
> [13] "min.d" ? ? "max.d" ? ? "temp.d" ?
?"slat" ? ? ?"slong" ? ? "spp"
> [19] "number" ? ?"weight" ? ?"elat" ? ?
?"elong"
>
> Each survey unit generates one set record, denoted by a 5 in column
"record.t". Each species identified in this particular survey unit
generates an additional set record, denoted by a 6.
>
>> unique(RV09$record.t)
> [1] 5 6
>
> Each survey unit are identified by a specific "trip" and
"set" number, so if there is a 5 record type with no associated 6
records, it means that no species were observed in that survey unit. I would
like to be able to select all and only these survey units, which represent my
zeros.
>
> So as an exemple, in this trip number 913, set 1, 3, and 4 would be part of
my "zeros" data.frame as they appear with no record.t 6, such that no
species were observed in this survey unit.
>
>> head(RV09)
> ? record.t trip set month stratum NAFO unit.area time dur.set distance
> 585 ? ? ? ?5 ?913 ? 1 ? ?10 ? ? 351 ? 3O ? ? ? R31 1044 ? ? ?17 ? ? ? ?9
> 586 ? ? ? ?5 ?913 ? 2 ? ?10 ? ? 351 ? 3O ? ? ? R31 1440 ? ? ?17 ? ? ? ?9
> 587 ? ? ? ?6 ?913 ? 2 ? ?10 ? ? 351 ? 3O ? ? ? R31 1440 ? ? ?17 ? ? ? ?9
> 588 ? ? ? ?5 ?913 ? 3 ? ?10 ? ? 340 ? 3O ? ? ? Q31 1800 ? ? ?18 ? ? ? ?9
> 589 ? ? ? ?5 ?913 ? 4 ? ?10 ? ? 340 ? 3O ? ? ? Q32 2142 ? ? ?17 ? ? ? ?9
>
> Any tips on how extract this "zero" data.frame in R?
> Thank you very much in advance!
>
> Best,
> ~Aurelie
>
>
> Aurelie Cosandey-Godin
> Ph.D. student, Department of Biology
> Industrial Graduate Fellow, WWF-Canada
> Dalhousie University | Email: godina at dal.ca
>
>
> ? ? ? ?[[alternative HTML version deleted]]
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>

Does this work?

library('plyr')
# Function to return a data frame if it has one row, else return NULL:
f <- function(d) if(nrow(d) == 1L) d else NULL
> ddply(RV09, .(set, month), f)
  record.t trip set month stratum NAFO unit.area time dur.set distance
1        5  913   1    10     351   3O       R31 1044      17        9
2        5  913   3    10     340   3O       Q31 1800      18        9
3        5  913   4    10     340   3O       Q32 2142      17        9

ddply() is an apply-like function that takes a data frame as input and
a data frame as output (hence the dd). The first argument is the data
frame name, the second argument the set of grouping variables and the
third is the function to be called (in this application).

HTH,
Dennis

On Tue, Nov 1, 2011 at 10:16 AM, Aurelie Cosandey Godin <godina at dal.ca>
wrote:
> Dear list,
>
> After reading different mails, blogs, and tried a few different codes
without any success, I am asking your help!
> I have the following data frame where each row represent a survey unit with
the following variables:
>
>> names(RV09)
> ?[1] "record.t" ?"trip" ? ? ?"set" ? ? ?
"month" ? ? "stratum" ? "NAFO"
> ?[7] "unit.area" "time" ? ? ?"dur.set" ?
"distance" ?"operation" "mean.d"
> [13] "min.d" ? ? "max.d" ? ? "temp.d" ?
?"slat" ? ? ?"slong" ? ? "spp"
> [19] "number" ? ?"weight" ? ?"elat" ? ?
?"elong"
>
> Each survey unit generates one set record, denoted by a 5 in column
"record.t". Each species identified in this particular survey unit
generates an additional set record, denoted by a 6.
>
>> unique(RV09$record.t)
> [1] 5 6
>
> Each survey unit are identified by a specific "trip" and
"set" number, so if there is a 5 record type with no associated 6
records, it means that no species were observed in that survey unit. I would
like to be able to select all and only these survey units, which represent my
zeros.
>
> So as an exemple, in this trip number 913, set 1, 3, and 4 would be part of
my "zeros" data.frame as they appear with no record.t 6, such that no
species were observed in this survey unit.
>
>> head(RV09)
> ? record.t trip set month stratum NAFO unit.area time dur.set distance
> 585 ? ? ? ?5 ?913 ? 1 ? ?10 ? ? 351 ? 3O ? ? ? R31 1044 ? ? ?17 ? ? ? ?9
> 586 ? ? ? ?5 ?913 ? 2 ? ?10 ? ? 351 ? 3O ? ? ? R31 1440 ? ? ?17 ? ? ? ?9
> 587 ? ? ? ?6 ?913 ? 2 ? ?10 ? ? 351 ? 3O ? ? ? R31 1440 ? ? ?17 ? ? ? ?9
> 588 ? ? ? ?5 ?913 ? 3 ? ?10 ? ? 340 ? 3O ? ? ? Q31 1800 ? ? ?18 ? ? ? ?9
> 589 ? ? ? ?5 ?913 ? 4 ? ?10 ? ? 340 ? 3O ? ? ? Q32 2142 ? ? ?17 ? ? ? ?9
>
> Any tips on how extract this "zero" data.frame in R?
> Thank you very much in advance!
>
> Best,
> ~Aurelie
>
>
> Aurelie Cosandey-Godin
> Ph.D. student, Department of Biology
> Industrial Graduate Fellow, WWF-Canada
> Dalhousie University | Email: godina at dal.ca
>
>
> ? ? ? ?[[alternative HTML version deleted]]
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>