VMI - incvars - missing data visualization

[mworni]

Ricardo - I started to play with VMI. I got some code to insert for the script. See page 3 of the "VIM imputation GUI" document I saved on the shared gdrive folder for the code in orginal.

incvars <- c(paste(c("age_groups", "diabetes", "alcoholuse", "chronicpulmonary", "decreasedfunction", "acutepulmonary", "chroniccardiac", "acutecardiac", "hypertension", "pvd", "renal", "acutecognitive", "cerebrovascular", "preopinfectedwound", "chronicsteroids", "preopalbumincateg", "bleedingdisorder", "totalwrvuquartile", "incisionalwoundclass", "optimequartile", "anesthesiaonlytimequartile", "ascites", "esophvarices")))
nsqip.data[, incvars] <- log10(nsqip.data[, incvars] + 1)
Fehler in [.data.frame(nsqip.data, , incvars) :
undefined columns selected

I know that the mistake comes up because of "age_groups" because when I take this variable out, I don't get the error message anymore - but I don't know why this happens. I tried to reclass age_groups to missing - but this didn't help.

Any idea?

Btw. same happens here when I add preopsepsisclass to the incvars command beldow.

CrossTable(preopsepsisclass, missing.include=TRUE)
incvars <- c(paste(c("bmiclass", "bmi", "highasa", "residentpresent", "preophospitallos")))
nsqip.data[, incvars] <- log10(nsqip.data[, incvars] + 1)
aggr(nsqip.data[, incvars], numbers=TRUE, prop=c(TRUE, FALSE))

[rpietro]

all that message means is that you don't have a column called age_groups --
if you do names(nsqip.data) you will see that this is true. not sure where
that variable is. there are other variables in that column that also aren't
in your data set

btw, would comment the stuff you don't use all the time, like the
install.package stuff -- takes centuries to run

On Fri, Jun 29, 2012 at 7:35 AM, mworni <
reply@reply.github.com

wrote:

Ricardo - I started to play with VMI. I got some code to insert for the
script. See page 3 of the "VIM imputation GUI" document I saved on the
shared gdrive folder for the code in orginal.

incvars <- c(paste(c("age_groups", "diabetes", "alcoholuse",
"chronicpulmonary", "decreasedfunction", "acutepulmonary",
"chroniccardiac", "acutecardiac", "hypertension", "pvd", "renal",
"acutecognitive", "cerebrovascular", "preopinfectedwound",
"chronicsteroids", "preopalbumincateg", "bleedingdisorder",
"totalwrvuquartile", "incisionalwoundclass", "optimequartile",
"anesthesiaonlytimequartile", "ascites", "esophvarices")))
nsqip.data[, incvars] <- log10(nsqip.data[, incvars] + 1)
Fehler in [.data.frame(nsqip.data, , incvars) :
undefined columns selected

I know that the mistake comes up because of "age_groups" because when I
take this variable out, I don't get the error message anymore - but I don't
know why this happens. I tried to reclass age_groups to missing - but this
didn't help.

Any idea?

Btw. same happens here when I add preopsepsisclass to the incvars command
beldow.

CrossTable(preopsepsisclass, missing.include=TRUE)
incvars <- c(paste(c("bmiclass", "bmi", "highasa", "residentpresent",
"preophospitallos")))
nsqip.data[, incvars] <- log10(nsqip.data[, incvars] + 1)
aggr(nsqip.data[, incvars], numbers=TRUE, prop=c(TRUE, FALSE))

---

Reply to this email directly or view it on GitHub:
#22

[mworni]

Don't understand - when I use the

CrossTable(age_groups, missing.include=TRUE)

Cell Contents
N
N / Table Total
-------------------------

Total Observations in Table: 100829

        | 18-39 years | 40-49 years | 50-59 years | 60-69 years | 70-79

years |

|-------------|-------------|-------------|-------------|-------------|
| 39060 | 16298 | 15837 | 12312 |
9149 |
| 0.387 | 0.162 | 0.157 | 0.122 |
0.091 |

|-------------|-------------|-------------|-------------|-------------|

        | 80-89 years |   90+ years |
        |-------------|-------------|
        |        6890 |        1283 |
        |       0.068 |       0.013 |
        |-------------|-------------|

I get a result - but when I use it in the incvars command - it doesn't
work. The age_groups variable does exist.

On Sun, Jul 1, 2012 at 6:36 AM, Ricardo Pietrobon <
reply@reply.github.com

wrote:

all that message means is that you don't have a column called age_groups --
if you do names(nsqip.data) you will see that this is true. not sure where
that variable is. there are other variables in that column that also aren't
in your data set

btw, would comment the stuff you don't use all the time, like the
install.package stuff -- takes centuries to run

On Fri, Jun 29, 2012 at 7:35 AM, mworni <
reply@reply.github.com

wrote:

Ricardo - I started to play with VMI. I got some code to insert for the
script. See page 3 of the "VIM imputation GUI" document I saved on the
shared gdrive folder for the code in orginal.

incvars <- c(paste(c("age_groups", "diabetes", "alcoholuse",
"chronicpulmonary", "decreasedfunction", "acutepulmonary",
"chroniccardiac", "acutecardiac", "hypertension", "pvd", "renal",
"acutecognitive", "cerebrovascular", "preopinfectedwound",
"chronicsteroids", "preopalbumincateg", "bleedingdisorder",
"totalwrvuquartile", "incisionalwoundclass", "optimequartile",
"anesthesiaonlytimequartile", "ascites", "esophvarices")))
nsqip.data[, incvars] <- log10(nsqip.data[, incvars] + 1)
Fehler in [.data.frame(nsqip.data, , incvars) :
undefined columns selected

I know that the mistake comes up because of "age_groups" because when I
take this variable out, I don't get the error message anymore - but I
don't
know why this happens. I tried to reclass age_groups to missing - but
this
didn't help.

Any idea?

Btw. same happens here when I add preopsepsisclass to the incvars command
beldow.

CrossTable(preopsepsisclass, missing.include=TRUE)
incvars <- c(paste(c("bmiclass", "bmi", "highasa", "residentpresent",
"preophospitallos")))
nsqip.data[, incvars] <- log10(nsqip.data[, incvars] + 1)
aggr(nsqip.data[, incvars], numbers=TRUE, prop=c(TRUE, FALSE))

---

Reply to this email directly or view it on GitHub:
#22

---

Reply to this email directly or view it on GitHub:

#22 (comment)

[mworni]

just realized that when I was creating the subset even after recoding my
variables that the created age_group variable did not arrive in the new
dataset.

Is it possible that age_group is not really attached to the nsqip.data
dataset?

On Sun, Jul 1, 2012 at 10:09 AM, Mathias Worni mathiasworni@gmail.comwrote:

Don't understand - when I use the

CrossTable(age_groups, missing.include=TRUE)

Cell Contents
N
N / Table Total
-------------------------

Total Observations in Table: 100829

        | 18-39 years | 40-49 years | 50-59 years | 60-69 years |

70-79 years |

|-------------|-------------|-------------|-------------|-------------|
| 39060 | 16298 | 15837 | 12312 |
9149 |
| 0.387 | 0.162 | 0.157 | 0.122 |
0.091 |

|-------------|-------------|-------------|-------------|-------------|

        | 80-89 years |   90+ years |
        |-------------|-------------|
        |        6890 |        1283 |
        |       0.068 |       0.013 |
        |-------------|-------------|

I get a result - but when I use it in the incvars command - it doesn't
work. The age_groups variable does exist.

On Sun, Jul 1, 2012 at 6:36 AM, Ricardo Pietrobon <
reply@reply.github.com

wrote:

all that message means is that you don't have a column called age_groups

if you do names(nsqip.data) you will see that this is true. not sure where
that variable is. there are other variables in that column that also
aren't
in your data set

btw, would comment the stuff you don't use all the time, like the
install.package stuff -- takes centuries to run

On Fri, Jun 29, 2012 at 7:35 AM, mworni <
reply@reply.github.com

wrote:

Ricardo - I started to play with VMI. I got some code to insert for the
script. See page 3 of the "VIM imputation GUI" document I saved on the
shared gdrive folder for the code in orginal.

incvars <- c(paste(c("age_groups", "diabetes", "alcoholuse",
"chronicpulmonary", "decreasedfunction", "acutepulmonary",
"chroniccardiac", "acutecardiac", "hypertension", "pvd", "renal",
"acutecognitive", "cerebrovascular", "preopinfectedwound",
"chronicsteroids", "preopalbumincateg", "bleedingdisorder",
"totalwrvuquartile", "incisionalwoundclass", "optimequartile",
"anesthesiaonlytimequartile", "ascites", "esophvarices")))
nsqip.data[, incvars] <- log10(nsqip.data[, incvars] + 1)
Fehler in [.data.frame(nsqip.data, , incvars) :
undefined columns selected

I know that the mistake comes up because of "age_groups" because when I
take this variable out, I don't get the error message anymore - but I
don't
know why this happens. I tried to reclass age_groups to missing - but
this
didn't help.

Any idea?

Btw. same happens here when I add preopsepsisclass to the incvars
command
beldow.

CrossTable(preopsepsisclass, missing.include=TRUE)
incvars <- c(paste(c("bmiclass", "bmi", "highasa", "residentpresent",
"preophospitallos")))
nsqip.data[, incvars] <- log10(nsqip.data[, incvars] + 1)
aggr(nsqip.data[, incvars], numbers=TRUE, prop=c(TRUE, FALSE))

---

Reply to this email directly or view it on GitHub:
#22

---

Reply to this email directly or view it on GitHub:

#22 (comment)

[rpietro]

so, i found that the variable wasn't there through names(nsqip.data) and
View(nsqip.data) and so the problem was not with the subset since it wasn't
there from the beginning. most important thing now is to track where the
variable is and where it was lost. would start by opening the original
(very first) dataset in a spreadsheet and checking

On Sun, Jul 1, 2012 at 4:23 AM, mworni <
reply@reply.github.com

wrote:

just realized that when I was creating the subset even after recoding my
variables that the created age_group variable did not arrive in the new
dataset.

Is it possible that age_group is not really attached to the nsqip.data
dataset?

On Sun, Jul 1, 2012 at 10:09 AM, Mathias Worni <mathiasworni@gmail.com

wrote:

Don't understand - when I use the

CrossTable(age_groups, missing.include=TRUE)

Cell Contents
N
N / Table Total
-------------------------

Total Observations in Table: 100829

        | 18-39 years | 40-49 years | 50-59 years | 60-69 years |

70-79 years |

|-------------|-------------|-------------|-------------|-------------|
| 39060 | 16298 | 15837 | 12312 |
9149 |
| 0.387 | 0.162 | 0.157 | 0.122 |
0.091 |

|-------------|-------------|-------------|-------------|-------------|

        | 80-89 years |   90+ years |
        |-------------|-------------|
        |        6890 |        1283 |
        |       0.068 |       0.013 |
        |-------------|-------------|

I get a result - but when I use it in the incvars command - it doesn't
work. The age_groups variable does exist.

On Sun, Jul 1, 2012 at 6:36 AM, Ricardo Pietrobon <
reply@reply.github.com

wrote:

all that message means is that you don't have a column called age_groups

if you do names(nsqip.data) you will see that this is true. not sure
where
that variable is. there are other variables in that column that also
aren't
in your data set

btw, would comment the stuff you don't use all the time, like the
install.package stuff -- takes centuries to run

On Fri, Jun 29, 2012 at 7:35 AM, mworni <
reply@reply.github.com

wrote:

Ricardo - I started to play with VMI. I got some code to insert for
the
script. See page 3 of the "VIM imputation GUI" document I saved on the
shared gdrive folder for the code in orginal.

incvars <- c(paste(c("age_groups", "diabetes", "alcoholuse",
"chronicpulmonary", "decreasedfunction", "acutepulmonary",
"chroniccardiac", "acutecardiac", "hypertension", "pvd", "renal",
"acutecognitive", "cerebrovascular", "preopinfectedwound",
"chronicsteroids", "preopalbumincateg", "bleedingdisorder",
"totalwrvuquartile", "incisionalwoundclass", "optimequartile",
"anesthesiaonlytimequartile", "ascites", "esophvarices")))
nsqip.data[, incvars] <- log10(nsqip.data[, incvars] + 1)
Fehler in [.data.frame(nsqip.data, , incvars) :
undefined columns selected

I know that the mistake comes up because of "age_groups" because when
I
take this variable out, I don't get the error message anymore - but I
don't
know why this happens. I tried to reclass age_groups to missing - but
this
didn't help.

Any idea?

Btw. same happens here when I add preopsepsisclass to the incvars
command
beldow.

CrossTable(preopsepsisclass, missing.include=TRUE)
incvars <- c(paste(c("bmiclass", "bmi", "highasa", "residentpresent",
"preophospitallos")))
nsqip.data[, incvars] <- log10(nsqip.data[, incvars] + 1)
aggr(nsqip.data[, incvars], numbers=TRUE, prop=c(TRUE, FALSE))

---

Reply to this email directly or view it on GitHub:
#22

---

Reply to this email directly or view it on GitHub:

#22 (comment)

---

Reply to this email directly or view it on GitHub:

#22 (comment)

[mworni]

I created this variable in the "variable recoding" section of the script.
It does not exist in the very original dataset.

On Sun, Jul 1, 2012 at 1:35 PM, Ricardo Pietrobon <
reply@reply.github.com

wrote:

so, i found that the variable wasn't there through names(nsqip.data) and
View(nsqip.data) and so the problem was not with the subset since it wasn't
there from the beginning. most important thing now is to track where the
variable is and where it was lost. would start by opening the original
(very first) dataset in a spreadsheet and checking

On Sun, Jul 1, 2012 at 4:23 AM, mworni <
reply@reply.github.com

wrote:

just realized that when I was creating the subset even after recoding my
variables that the created age_group variable did not arrive in the new
dataset.

Is it possible that age_group is not really attached to the nsqip.data
dataset?

On Sun, Jul 1, 2012 at 10:09 AM, Mathias Worni <mathiasworni@gmail.com

wrote:

Don't understand - when I use the

CrossTable(age_groups, missing.include=TRUE)

Cell Contents
N
N / Table Total
-------------------------

Total Observations in Table: 100829

        | 18-39 years | 40-49 years | 50-59 years | 60-69 years |

70-79 years |

|-------------|-------------|-------------|-------------|-------------|
| 39060 | 16298 | 15837 | 12312 |
9149 |
| 0.387 | 0.162 | 0.157 | 0.122 |
0.091 |

|-------------|-------------|-------------|-------------|-------------|

        | 80-89 years |   90+ years |
        |-------------|-------------|
        |        6890 |        1283 |
        |       0.068 |       0.013 |
        |-------------|-------------|

I get a result - but when I use it in the incvars command - it doesn't
work. The age_groups variable does exist.

On Sun, Jul 1, 2012 at 6:36 AM, Ricardo Pietrobon <
reply@reply.github.com

wrote:

all that message means is that you don't have a column called

age_groups

if you do names(nsqip.data) you will see that this is true. not sure
where
that variable is. there are other variables in that column that also
aren't
in your data set

btw, would comment the stuff you don't use all the time, like the
install.package stuff -- takes centuries to run

On Fri, Jun 29, 2012 at 7:35 AM, mworni <
reply@reply.github.com

wrote:

Ricardo - I started to play with VMI. I got some code to insert for
the
script. See page 3 of the "VIM imputation GUI" document I saved on
the
shared gdrive folder for the code in orginal.

incvars <- c(paste(c("age_groups", "diabetes", "alcoholuse",
"chronicpulmonary", "decreasedfunction", "acutepulmonary",
"chroniccardiac", "acutecardiac", "hypertension", "pvd", "renal",
"acutecognitive", "cerebrovascular", "preopinfectedwound",
"chronicsteroids", "preopalbumincateg", "bleedingdisorder",
"totalwrvuquartile", "incisionalwoundclass", "optimequartile",
"anesthesiaonlytimequartile", "ascites", "esophvarices")))
nsqip.data[, incvars] <- log10(nsqip.data[, incvars] + 1)
Fehler in [.data.frame(nsqip.data, , incvars) :
undefined columns selected

I know that the mistake comes up because of "age_groups" because
when
I
take this variable out, I don't get the error message anymore - but
I
don't
know why this happens. I tried to reclass age_groups to missing -
but
this
didn't help.

Any idea?

Btw. same happens here when I add preopsepsisclass to the incvars
command
beldow.

CrossTable(preopsepsisclass, missing.include=TRUE)
incvars <- c(paste(c("bmiclass", "bmi", "highasa",
"residentpresent",
"preophospitallos")))
nsqip.data[, incvars] <- log10(nsqip.data[, incvars] + 1)
aggr(nsqip.data[, incvars], numbers=TRUE, prop=c(TRUE, FALSE))

---

Reply to this email directly or view it on GitHub:
#22

---

Reply to this email directly or view it on GitHub:

#22 (comment)

---

Reply to this email directly or view it on GitHub:

#22 (comment)

---

Reply to this email directly or view it on GitHub:

#22 (comment)

[rpietro]

this is getting hard -- i can't organize the file because we are working at
the same time, and withe file as is it is taking me forever because there
is a bunch of stuff that needs to be distributed across different files.
here is what i suggest:

1. will wait for you to stop working at night and then i will come in and
   re-organize everything
2. at the level where we are, it might be easier for you to learn github
   since then we won't have any other file conflicts. but this will take time,
   and so will leave that decision up to you. i'm fine with whatever you might
   decide

On Sun, Jul 1, 2012 at 7:39 AM, mworni <
reply@reply.github.com

wrote:

I created this variable in the "variable recoding" section of the script.
It does not exist in the very original dataset.

On Sun, Jul 1, 2012 at 1:35 PM, Ricardo Pietrobon <
reply@reply.github.com

wrote:

so, i found that the variable wasn't there through names(nsqip.data) and
View(nsqip.data) and so the problem was not with the subset since it
wasn't
there from the beginning. most important thing now is to track where the
variable is and where it was lost. would start by opening the original
(very first) dataset in a spreadsheet and checking

On Sun, Jul 1, 2012 at 4:23 AM, mworni <
reply@reply.github.com

wrote:

just realized that when I was creating the subset even after recoding
my
variables that the created age_group variable did not arrive in the new
dataset.

Is it possible that age_group is not really attached to the nsqip.data
dataset?

On Sun, Jul 1, 2012 at 10:09 AM, Mathias Worni <mathiasworni@gmail.com

wrote:

Don't understand - when I use the

CrossTable(age_groups, missing.include=TRUE)

Cell Contents
N
N / Table Total
-------------------------

Total Observations in Table: 100829

        | 18-39 years | 40-49 years | 50-59 years | 60-69 years |

70-79 years |

|-------------|-------------|-------------|-------------|-------------|
| 39060 | 16298 | 15837 | 12312 |
9149 |
| 0.387 | 0.162 | 0.157 | 0.122 |
0.091 |

|-------------|-------------|-------------|-------------|-------------|

        | 80-89 years |   90+ years |
        |-------------|-------------|
        |        6890 |        1283 |
        |       0.068 |       0.013 |
        |-------------|-------------|

I get a result - but when I use it in the incvars command - it
doesn't
work. The age_groups variable does exist.

On Sun, Jul 1, 2012 at 6:36 AM, Ricardo Pietrobon <
reply@reply.github.com

wrote:

all that message means is that you don't have a column called

age_groups

if you do names(nsqip.data) you will see that this is true. not sure
where
that variable is. there are other variables in that column that also
aren't
in your data set

btw, would comment the stuff you don't use all the time, like the
install.package stuff -- takes centuries to run

On Fri, Jun 29, 2012 at 7:35 AM, mworni <
reply@reply.github.com

wrote:

Ricardo - I started to play with VMI. I got some code to insert
for
the
script. See page 3 of the "VIM imputation GUI" document I saved on
the
shared gdrive folder for the code in orginal.

incvars <- c(paste(c("age_groups", "diabetes", "alcoholuse",
"chronicpulmonary", "decreasedfunction", "acutepulmonary",
"chroniccardiac", "acutecardiac", "hypertension", "pvd", "renal",
"acutecognitive", "cerebrovascular", "preopinfectedwound",
"chronicsteroids", "preopalbumincateg", "bleedingdisorder",
"totalwrvuquartile", "incisionalwoundclass", "optimequartile",
"anesthesiaonlytimequartile", "ascites", "esophvarices")))
nsqip.data[, incvars] <- log10(nsqip.data[, incvars] + 1)
Fehler in [.data.frame(nsqip.data, , incvars) :
undefined columns selected

I know that the mistake comes up because of "age_groups" because
when
I
take this variable out, I don't get the error message anymore -
but
I
don't
know why this happens. I tried to reclass age_groups to missing -
but
this
didn't help.

Any idea?

Btw. same happens here when I add preopsepsisclass to the incvars
command
beldow.

CrossTable(preopsepsisclass, missing.include=TRUE)
incvars <- c(paste(c("bmiclass", "bmi", "highasa",
"residentpresent",
"preophospitallos")))
nsqip.data[, incvars] <- log10(nsqip.data[, incvars] + 1)
aggr(nsqip.data[, incvars], numbers=TRUE, prop=c(TRUE, FALSE))

---

Reply to this email directly or view it on GitHub:
#22

---

Reply to this email directly or view it on GitHub:

#22 (comment)

---

Reply to this email directly or view it on GitHub:

#22 (comment)

---

Reply to this email directly or view it on GitHub:

#22 (comment)

---

Reply to this email directly or view it on GitHub:

#22 (comment)

[rpietro]

thought a bit more about this. if i change the file structure now, with you
coming back to the clinic, we risk confusing you. so, let's leave as is.
let me know if the changes we discussed today fixed the issue. it seems
like the real problem was really just the missing variable, which we might
have fixed through the detaching of the data

On Sun, Jul 1, 2012 at 9:34 AM, Ricardo Pietrobon pietr007@gmail.comwrote:

this is getting hard -- i can't organize the file because we are working
at the same time, and withe file as is it is taking me forever because
there is a bunch of stuff that needs to be distributed across different
files. here is what i suggest:

1. will wait for you to stop working at night and then i will come in and
   re-organize everything
2. at the level where we are, it might be easier for you to learn github
   since then we won't have any other file conflicts. but this will take time,
   and so will leave that decision up to you. i'm fine with whatever you might
   decide

On Sun, Jul 1, 2012 at 7:39 AM, mworni <
reply@reply.github.com

wrote:

I created this variable in the "variable recoding" section of the script.
It does not exist in the very original dataset.

On Sun, Jul 1, 2012 at 1:35 PM, Ricardo Pietrobon <
reply@reply.github.com

wrote:

so, i found that the variable wasn't there through names(nsqip.data) and
View(nsqip.data) and so the problem was not with the subset since it
wasn't
there from the beginning. most important thing now is to track where the
variable is and where it was lost. would start by opening the original
(very first) dataset in a spreadsheet and checking

On Sun, Jul 1, 2012 at 4:23 AM, mworni <
reply@reply.github.com

wrote:

just realized that when I was creating the subset even after recoding
my
variables that the created age_group variable did not arrive in the
new
dataset.

Is it possible that age_group is not really attached to the nsqip.data
dataset?

On Sun, Jul 1, 2012 at 10:09 AM, Mathias Worni <
mathiasworni@gmail.com

wrote:

Don't understand - when I use the

CrossTable(age_groups, missing.include=TRUE)

Cell Contents
N
N / Table Total
-------------------------

Total Observations in Table: 100829

        | 18-39 years | 40-49 years | 50-59 years | 60-69 years

|
|

|-------------|-------------|-------------|-------------|-------------|
| 39060 | 16298 | 15837 | 12312
|
9149 |
| 0.387 | 0.162 | 0.157 | 0.122
|
0.091 |

|-------------|-------------|-------------|-------------|-------------|

        | 80-89 years |   90+ years |
        |-------------|-------------|
        |        6890 |        1283 |
        |       0.068 |       0.013 |
        |-------------|-------------|

I get a result - but when I use it in the incvars command - it
doesn't
work. The age_groups variable does exist.

On Sun, Jul 1, 2012 at 6:36 AM, Ricardo Pietrobon <
reply@reply.github.com

wrote:

all that message means is that you don't have a column called

age_groups

if you do names(nsqip.data) you will see that this is true. not
sure
where
that variable is. there are other variables in that column that
also
aren't
in your data set

btw, would comment the stuff you don't use all the time, like the
install.package stuff -- takes centuries to run

On Fri, Jun 29, 2012 at 7:35 AM, mworni <
reply@reply.github.com

wrote:

Ricardo - I started to play with VMI. I got some code to insert
for
the
script. See page 3 of the "VIM imputation GUI" document I saved
on
the
shared gdrive folder for the code in orginal.

incvars <- c(paste(c("age_groups", "diabetes", "alcoholuse",
"chronicpulmonary", "decreasedfunction", "acutepulmonary",
"chroniccardiac", "acutecardiac", "hypertension", "pvd", "renal",
"acutecognitive", "cerebrovascular", "preopinfectedwound",
"chronicsteroids", "preopalbumincateg", "bleedingdisorder",
"totalwrvuquartile", "incisionalwoundclass", "optimequartile",
"anesthesiaonlytimequartile", "ascites", "esophvarices")))
nsqip.data[, incvars] <- log10(nsqip.data[, incvars] + 1)
Fehler in [.data.frame(nsqip.data, , incvars) :
undefined columns selected

I know that the mistake comes up because of "age_groups" because
when
I
take this variable out, I don't get the error message anymore -
but
I
don't
know why this happens. I tried to reclass age_groups to missing -
but
this
didn't help.

Any idea?

Btw. same happens here when I add preopsepsisclass to the incvars
command
beldow.

CrossTable(preopsepsisclass, missing.include=TRUE)
incvars <- c(paste(c("bmiclass", "bmi", "highasa",
"residentpresent",
"preophospitallos")))
nsqip.data[, incvars] <- log10(nsqip.data[, incvars] + 1)
aggr(nsqip.data[, incvars], numbers=TRUE, prop=c(TRUE, FALSE))

---

Reply to this email directly or view it on GitHub:
#22

---

Reply to this email directly or view it on GitHub:

#22 (comment)

---

Reply to this email directly or view it on GitHub:

#22 (comment)

---

Reply to this email directly or view it on GitHub:

#22 (comment)

---

Reply to this email directly or view it on GitHub:

#22 (comment)