I'm working, in RStudio, with data for patients that are either normal, have Crohn's disease, or ulcerative colitis. Now, the data is structured in such a way that patient information is in a separate data frame (called sampleInfo), and the data I want to use for analysis is in a different data frame (called expressionData). For my analysis, I would like to remove the patients that are 'normal' from the dataset and only keep those with Crohn's disease or ulcerative colitis.
So, what I did was first run the following command to make a new data frame from sampleInfo containing all the patients (aka rows) with the normal disease state, using the following command:
bad_patients <- sampleInfo[sampleInfo$characteristics_ch1.3 == "disease state: normal", ]
bad_patients has a column called geoaccession, which contains the patient ID, which also corresponds with the column names for the same patient in expressionData.
I save the names of these IDs using
patient_names <- bad_patients$geo_accession.
Now, I want to remove the columns with these names from expressionData. I looked at a lot of different StackOverflow posts, as well as posts on the R help forum, and found two main ways, both of which I have tried. The first is done with the following command:
newDataFrame <- expressionData[ , !names(expressionData) %in% patient_names]
Though this method does produce a new matrix called newDataFrame, attempting to view this matrix in RStudio gives the following error:
Error in View : 'names' attribute [1] must be the same length as the vector [0]
I also tried a second subset method with the following command:
newDataFrame <- subset(expressionData, -patient_names)
which raises the error: Error in -patient_names : invalid argument to unary operator
I also tried this subset method by explicity typing out the columns I wanted to remove as follows:
newDataFrame <- subset(expressionData, -c('ID090190', ...) (where ... corresponds to the rest of the IDs) and got the same exact error.
Can someone tell me what I'm doing wrong, or how to work around this?
Couple of solutions:
Subsetting based on names
newDataFrame <- expressionData[!(names(expressionData) %in% patient_names)]
One problem with your attempt was that you hadn't wrapped the whole expression evaluated by ! in parentheses. As it was, you were looking for !names(expressionData) in patient_names. ! here would coerce names(expressionData) into a logical and likely return a vector full of FALSEs
I've subset with only one dimension (x[this] rather than x[,this]). You can do this with the columns of data frames because a data frame is a list of its columns. This subsetting method preserves the data.frame class of the returned object, whereas the two-dimensional subset will just return a vector if you select only one column. (Tibbles will return a tibble with both methods, which is one big advantage of tibbles)
Tidyverse solution: use dplyr::select with dplyr::all_of
newDataFrame <- dplyr::select(expressionData, -dplyr::all_of(patientnames))
Edit: Make sure your data really is a data.frame
If you're getting this error Error in UseMethod("select_") : no applicable method for 'select_' applied to an object of class "c('matrix', 'array', 'double', 'numeric')", it's because your data is a matrix, rather than a data frame. You may have inadvertently coerced it in processing.
Use as.data.frame to return to a data frame object, which will be compabtible with the methods above. If you wish to keep your data as a matrix, use colnames:
expressionData[ , !(colnames(expressionData) %in% patient_names)] to subset the columns.
If expressionData is a matrix, you'll need to subset the columns with colnames, rather than names. The names of a data.frame are identical to its colnames (because a df is a list of its columns), but the names of a matrix are the names of every element in the matrix, because a matrix is just an array with dimensionality. You'll want to check colnames(expressionData) to make sure that there are colnames to subset.
You might want to try:
newDataFrame <- expressionData[ , !colnames(expressionData) %in% patient_numbers]
names(expressionData) is NULL, hence your error; you want the column names
in your example, your list of sample names was called patient_numbers, not patient_names
Related
Instructions:*
Create an object called response that contains the name of the variable that will be considered the response variable.
Create a vector called predictors that contains the names of the variables that will be used as predictors.
Assignment Code:
# Create response object
response = 'price'
# Create predictors object
predictors = c("carat","cut","clarity","color","depth")
Feedback:
Now that you've identified what variables are of interest, you can subset the data to only include those columns.
Subsetting the Data
Context:
There are various ways to subset a data frame to only include specific columns. Here, you can use the objects response and predictors to indicate the names of the columns to keep. The function select(), from the {dplyr} package, is a good way to accomplish this task. It requires two arguments:
the name of the data frame being used
the column names to be included in the subset, or the objects that contain that information (response and predictors), separated by a comma
Instructions:
Use select() to create a subset of myData that contains only the columns of interest, and store this subset in an object called myData_subset.
Assignment Code:
# Subset the data
myData_subset %>% select(response,predictors)
Could anyone tell me where I am going wrong on subsetting? Is it that my objects were created incorrectly? Thanks a lot.
Try this: the assignment is lacking myData and Using an external vector in selections is ambiguous -> we could use all_of(..)
response = 'price'
predictors = c("carat","cut","clarity","color","depth")
myData_subset <- myData %>%
select(response,all_of(predictors))
I have n data frames, each corresponding to data from a city.
There are 3 variables per data frame and currently they are all factor variables.
I want to transform all of them into numeric variables.
I have started by creating a vector with the names of all the data frames in order to use in a for loop.
cities <- as.vector(objects())
for ( i in cities){
i <- as.data.frame(lapply(i, function(x) as.numeric(levels(x))[x]))
}
Although the code runs and there I get no error code, I don't see any changes to my data frames as all three variables remain factor variables.
The strangest thing is that when doing them one by one (as below) it works:
df <- as.data.frame(lapply(df, function(x) as.numeric(levels(x))[x]))
What you're essentially trying to do is modify the type of the field if it is a factor (to a numeric type). One approach using purrr would be:
library(purrr)
map(cities, ~ modify_if(., is.factor, as.numeric))
Note that modify() in itself is like lapply() but it doesn't change the underlying data structure of the objects you are modifying (in this case, dataframes). modify_if() simply takes a predicate as an additional argument.
for anyone who's interested in my question, I worked out the answer:
for ( i in cities){
assign(i, as.data.frame(lapply(get(i), function(x) as.numeric(levels(x))[x])))
}
I am simply trying to create a dataframe.
I read in data by doing:
>example <- read.csv(choose.files(), header=TRUE, sep=";")
The data contains 2 columns with 8736 rows plus a header.
I then simply want to combine this with the column of a dataframe with the same amount of rows (!) by doing:
>data_frame <- as.data.frame(example$x, example$y, otherdata$z)
It produces the following error
Warning message:
In as.data.frame.numeric(example$x, example$y, otherdata$z) :
'row.names' is not a character vector of length 8736 -- omitting it. Will be an error!
I have never had this problem before. It seems so easy to tackle but I cant help myself at the moment.
Overview
As long as the nrow(example) equals length(otherdata$z), use cbind.data.frame to combine columns into one data frame. An advantage with cbind.data.frame() is that there is no need to call the individual columns within example when binding them with otherdata$z.
# create a new data frame that adds the 'z' field from another source
df_example <- cbind.data.frame(example, otherdata$z)
I have a data frame consisting of five character variables which represent specific bacteria. I then have thousands of observations of each variable that all begin with the letter K. eg
x <- c(K0001,K0001,K0003,K0006)
y <- c(K0001,K0001,K0002,K0003)
z <- c(K0001,K0002,K0007,K0008)
r <- c(K0001,K0001,K0001,K0001)
o <- c(K0003,K0009,K0009,K0009)
I need to identify unique observations in the first column that don't appear in any of the remaining four columns. I have tried the approach suggested here which I think would work if I could create individual vectors using select ...
How to tell what is in one vector and not another?
but when I try to create a vector for analysis using the code ...
x <- select(data$x)
I get the error
Error in UseMethod("select_") :
no applicable method for 'select_' applied to an object of class "character
I have tried to mutate the vectors using as.factor and as.numeric but neither of these approaches work as the first gives an equivalent error as above, and as.numeric returns NAs.
Thanks in advance
The reference that you cited recommended using setdiff. The only thing that you need to do to apply that solution is to convert the four columns into one, so that it can be treated as a set. You can do that with unlist
setdiff(data$x, unlist(data[,2:5]))
"K0006"
Consider the following simulation snippet:
k <- 1:5
x <- seq(0,10,length.out = 100)
dsts <- lapply(1:length(k), function(i) cbind(x=x, distri=dchisq(x,k[i]),i) )
dsts <- do.call(rbind,dsts)
why does this code throws an error (dsts is matrix):
subset(dsts,i==1)
#Error in subset.matrix(dsts, i == 1) : object 'i' not found
Even this one:
colnames(dsts)[3] <- 'iii'
subset(dsts,iii==1)
But not this one (matrix coerced as dataframe):
subset(as.data.frame(dsts),i==1)
This one works either where x is already defined:
subset(dsts,x> 500)
The error occurs in subset.matrix() on this line:
else if (!is.logical(subset))
Is this a bug that should be reported to R Core?
The behavior you are describing is by design and is documented on the ?subset help page.
From the help page:
For data frames, the subset argument works on the rows. Note that subset will be evaluated in the data frame, so columns can be referred to (by name) as variables in the expression (see the examples).
In R, data.frames and matrices are very different types of objects. If this is causing a problem, you are probably using the wrong data structure for your data. Matrices are really only necessary if you meed matrix arithmetic. If you are thinking of your columns as different attributes for a row observations, then you should be storing your data in a data.frame in the first place. You could store all your values in a simple vector where every three values represent one observation, but that would also be a poor choice of data structure for your data. I'm not sure if you were trying to be more efficient by choosing a matrix but it seems like just the wrong choice.
A data.frame is stored as a named list while a matrix is stored as a dimensioned vector. A list can be used as an environment which makes it easy to evaluate variable names in that context. The biggest difference between the two is that data.frames can hold columns of different classes (numerics, characters, dates) while matrices can only hold values of exactly one data.type. You cannot always easily convert between the two without a loss of information.
Thinks like $ only work with data.frames as well.
dd <- data.frame(x=1:10)
dd$x
mm <- matrix(1:10, ncol=1, dimnames=list(NULL, "x"))
mm$x # Error
If you want to subset a matrix, you are better off using standard [ subsetting rather than the sub setting function.
dsts[ dsts[,"i"]==1, ]
This behavior has been a part of R for a very long time. Any changes to this behavior is likely to introduce breaking changes to existing code that relies on variables being evaluated in a certain context. I think the problem lies with whomever told you to use a matrix in the first place. Rather than cbind(), you should have used data.frame()