Reshape

Hello, helpeRs,

I've been trying, unsuccessfully, to change a dataframe from long to
wide format using reshape (the original). ?I would appreciate it if
someone could demonstrate the correct syntax. ?The script below will
create a toy example. ?The new wide data should have a column name
for each unique entry in the "fame" column. ? Under each column
should be either the appropriate weight or na, if there is no match.
Thus, the end product should have column names:

valley ?plot ? trt ? 18w ? 16iso ?12:0, etc.

Here is the toy script for the starting data in long form:

dat <- data.frame(fame = ?gl(4,1,10, labels = c( "18w", "16iso",
"12:0", "16w")), valley = gl(2,6,10, labels = c("H", "M")), plot =
gl(5, 2, 10), trt = gl(2,1,10, labels = c("e", "g")), weight =
1000*runif(10, 0, 1))

Thank you for your assistance.

Toby

Hi Toby,

I think this should work...

reshape(dat, v.names=c("weight"), idvar=c("valley", "plot", "trt"),
timevar="fame", direction="wide")

Michael


On 28 September 2010 12:17, Toby Gass <tobygass at warnercnr.colostate.edu> wrote:

Hello, helpeRs,

I've been trying, unsuccessfully, to change a dataframe from long to
wide format using reshape (the original). ?I would appreciate it if
someone could demonstrate the correct syntax. ?The script below will
create a toy example. ?The new wide data should have a column name
for each unique entry in the "fame" column. ? Under each column
should be either the appropriate weight or na, if there is no match.
Thus, the end product should have column names:

valley ?plot ? trt ? 18w ? 16iso ?12:0, etc.

Here is the toy script for the starting data in long form:

dat <- data.frame(fame = ?gl(4,1,10, labels = c( "18w", "16iso",
"12:0", "16w")), valley = gl(2,6,10, labels = c("H", "M")), plot =
gl(5, 2, 10), trt = gl(2,1,10, labels = c("e", "g")), weight =
1000*runif(10, 0, 1))

Thank you for your assistance.

Toby

That does work, thank you. ?I didn't understand that the "fame"
column would be the time varying column.

Toby

On 28 Sep 2010 at 12:47, Michael Bedward wrote:

Hi Toby,

I think this should work...

reshape(dat, v.names=c("weight"), idvar=c("valley", "plot", "trt"),
timevar="fame", direction="wide")

Michael


On 28 September 2010 12:17, Toby Gass <tobygass at warnercnr.colostate.edu> wrote:

Hello, helpeRs,

I've been trying, unsuccessfully, to change a dataframe from long to
wide format using reshape (the original). ?I would appreciate it if
someone could demonstrate the correct syntax. ?The script below will
create a toy example. ?The new wide data should have a column name
for each unique entry in the "fame" column. ? Under each column
should be either the appropriate weight or na, if there is no match.
Thus, the end product should have column names:

valley ?plot ? trt ? 18w ? 16iso ?12:0, etc.

Here is the toy script for the starting data in long form:

dat <- data.frame(fame = ?gl(4,1,10, labels = c( "18w", "16iso",
"12:0", "16w")), valley = gl(2,6,10, labels = c("H", "M")), plot =
gl(5, 2, 10), trt = gl(2,1,10, labels = c("e", "g")), weight =
1000*runif(10, 0, 1))

Thank you for your assistance.

Toby