Is it possible to use ggplot without first assigning names to the variables in the data.frame? My intended use is early exploration of large datasets, the kind that involves trying one form of a question or possibility then moving on to the next. I record short comments about what is being explored but seldom name the rows and columns of the one-time-use variables I derive.
To see the data, I make quick and dirty plots:
plot(matrix.name)
or more often
plot(x = array.name[1, , 1], y=array.name[ , , 1])
At times I’d like to use ggplot2’s features instead. The requisite as.data.frame(matrix.name) conversion is quick, but then is there a way to pass the necessary arguments to aes() without assigning row and column names? As best I have been able to research, aes requires variable names.
Thank you very much for your help. I will repeatedly have occasion to use any answers.
Related
I'm wondering what are the best practices to change a dataframe's column types ideally using tidy selection languages.
Ideally you would set the col types correctly up front when you import the data but that isn't always possible for various reasons.
So the next best pattern that I could identify is the below:
#random dataframe
df <- tibble(a_col=1:10,
b_col=letters[1:10],
c_col=seq.Date(ymd("2022-01-01"),by="day",length.out = 10))
My current favorite pattern involves using across() because I can use tidy selection verb to select variables that I want and then can "map" a formula to those.
# current favorite pattern
df<- df %>%
mutate(across(starts_with("a"),as.character))
Does anyone have any other favorite patterns or useful tricks here? It doesn't have to mutate. Often times I have to change the column types of dataframes with 100s of columns so it becomes quite tedious.
Yes this happens. Pain is where dates are in character format and if you once modify them and try to modify again (say in a mutate / summarise) there will be error.
In such a cases, change datatype only when you get to know what kind of data is there.
Select with names of columns id there is a sense in them
Check before applying the as.* if its already in that type with is.*
Applying it can be be by map / lapply / for loop, whatever is comfortable.
But it would be difficult to have a single approach for "all dataframes" as people try to name fields as per their choice or convenience.
Shared mine. Hope others help.
Closed. This question is opinion-based. It is not currently accepting answers.
Want to improve this question? Update the question so it can be answered with facts and citations by editing this post.
Closed 5 years ago.
Improve this question
I am learning R through tutorials, but I have difficulties in "how to read" R code, which in turn makes it difficult to write R code. For example:
dir.create(file.path("testdir2","testdir3"), recursive = TRUE)
vs
dup.names <- as.character(data.combined[which(duplicated(as.character(data.combined$name))), "name"])
While I know what these lines of code do, I cannot read or interpret the logic of each line of code. Whether I read left to right or right to left. What strategies should I use when reading/writing R code?
dup.names <- as.character(data.combined[which(duplicated(as.character(data.combined$name))), "name"])
Don't let lines of code like this ruin writing R code for you
I'm going to be honest here. The code is bad. And for many reasons.
Not a lot of people can read a line like this and intuitively know what the output is.
The point is you should not write lines of code that you don't understand. This is not Excel, you do not have but 1 single line to fit everything within. You have a whole deliciously large script, an empty canvas. Use that space to break your code into smaller bits that make a beautiful mosaic piece of art! Let's dive in~
Dissecting the code: Data Frames
Reading a line of code is like looking at a face for familiar features. You can read left to right, middle to out, whatever -- as long as you can lock onto something that is familiar.
Okay you see data.combined. You know (hope) it has rows and columns... because it's data!
You spot a $ in the code and you know it has to be a data.frame. This is because only lists and data.frames (which are really just lists) allow you to subset columns using $ followed by the column name. Subset-by the way- just means looking at a portion of the overall. In R, subsetting for data.frames and matrices can be done using single brackets[, within which you will see [row, column]. Thus if we type data.combined[1,2], it would give you the value in row 1 of column 2.
Now, if you knew that the name of column 2 was name you can use data.combined[1,"name"] to get the same output as data.combined$name[1]. Look back at that code:
dup.names <- as.character(data.combined[which(duplicated(as.character(data.combined$name))), "name"])
Okay, so now we see our eyes should be locked on data.combined[SOMETHING IS IN HERE?!]) and slowly be picking out data.combined[ ?ROW? , Oh the "name" column]. Cool.
Finding those ROW values!
which(duplicated(as.character(data.combined$name)))
Anytime you see the which function, it is just giving you locations. An example: For the logical vector a = c(1,2,2,1), which(a == 1) would give you 1 and 4, the location of 1s in a.
Now duplicated is simple too. duplicated(a) (which is just duplicated(c(1,2,2,1))) will give you back FALSE FALSE TRUE TRUE. If we ran which(duplicated(a)) it would return 3 and 4. Now here is a secret you will learn. If you have TRUES and FALSES, you don't need to use the which function! So maybe which was unnessary here. And also as.character... since duplicated works on numbers and strings.
What You Should Be Writing
Who am I to tell you how to write code? But here's my take.
Don't mix up ways of subsetting: use EITHER data.frame[,column] or data.frame$column...
The code could have been written a little bit more legibly as:
dupes <- duplicated(data.combined$name)
dupe.names <- data.combines$name[dupes]
or equally:
dupes <- duplicated(data.combined[,"name"])
dupe.names <- data.combined[dupes,"name"]
I know this was lengthy but I hope it helps.
An easier way to read any code is to break up their components.
dup.names <-
as.character(
data.combined[which(
duplicated(
as.character(
data.combined$name
)
)
), "name"]
)
For each of the functions - those parts with rounded brackets following them e.g. as.character() you can learn more about what they do and how they work by typing ?as.character in the console
Square brackets [] are use to subset data frames, which are stored in your environment (the box to the upper right if you're using R within RStudio contains your values as well as any defined functions). In this case, you can tell that data.combined is the name that has been given to such a data frame in this example (type ?data.frame to find out more about data frames).
"Unwrapping" long lines of code can be daunting at first. Start by breaking it down into parenthesis , brackets, and commas. Parenthesis directly tacked onto a word indicate a function, and any commas that lie within them (unless they are part of another nested function or bracket) separate arguments which contain parameters that modify the way the function behaves. We can reduce your 2nd line to an outer function as.character and its arguments:
dup.names <- as.character(argument_1)
Just from this, we know that dup.names will be assigned a value with the data type "character" off of a single argument.
Two functions in the first line, file.path() and dir.create(), contain a comma to denote two arguments. Arguments can either be a single value or specified with an equal sign. In this case, the output of file.path happens to perform as argument #1 of dir.create().
file.path(argument_1,argument_2)
dir.create(argument_1,argument_2)
Brackets are a way of subsetting data frames, with the general notation of dataframe_object[row,column]. Within your second line is a dataframe object, data.combined. You know it's a dataframe object because of the brackets directly tacked onto it, and knowing this allows you to that any functions internal to this are contributing to subsetting this data frame.
data.combined[row, column]
So from there, we can see that the internal functions within this bracket will produce an output that specifies the rows of data.combined that will contribute to the subset, and that only columns with name "name" will be selected.
Use the help function to start to unpack these lines by discovering what each function does, and what it's arguments are.
I want to sort a data.frame by multiple columns, ideally using base R without any external packages (though if necessary, so be it). Having read How to sort a dataframe by column(s)?, I know I can accomplish this with the order() function as long as I either:
Know the explicit names of each of the columns.
Have a separate object representing each individual column by which to sort.
But what if I only have one vector containing multiple column names, of length that's unknown in advance?
Say the vector is called sortnames.
data[order(data[, sortnames]), ] won't work, because order() treats that as a single sorting argument.
data[order(data[, sortnames[1]], data[, sortnames[2]], ...), ] will work if and only if I specify the exact correct number of sortname values, which I won't know in advance.
Things I've looked at but not been totally happy with:
eval(parse(text=paste("data[with(data, order(", paste(sortnames, collapse=","), ")), ]"))). Maybe this is fine, but I've seen plenty of hate for using eval(), so asking for alternatives seemed worthwhile.
I may be able to use the Deducer library to do this with sortData(), but like I said, I'd rather avoid using external packages.
If I'm being too stubborn about not using external packages, let me know. I'll get over it. All ideas appreciated in advance!
You can use do.call:
data<-data.frame(a=rnorm(10),b=rnorm(10))
data<-data.frame(a=rnorm(10),b=rnorm(10),c=rnorm(10))
sortnames <- c("a", "b")
data[do.call("order", data[sortnames]), ]
This trick is useful when you want to pass multiple arguments to a function and these arguments are in convenient named list.
When doing sequencing, I normally apply TraMineR's seqdef function on a dataset to generate a single sequence object:
sequence_object <- seqdef(data)
However, let's say I want to loop through a dataframe and generate 1 sequence object per every chunk of 10 columns. Then I would do something like this:
colpicks <- seq(10,1000,by=10)
mapply(function(start,stop) seqdef(df[,start:stop]), colpicks-9, colpicks)
Now, I want to store these objects in some suitable manner. Two questions:
What is the most suitable way of storing (or maybe just automatically naming) 100 objects, so that I can easily loop through each of them at a later point?
How can I modify my code above so that it stores the data per your answer to (1)?
"Most suitable" is completely subjective and dependent on your goal.
I'm assuming this question is related to your previous question, and thus I would suggest setting the simplify argument of mapply to FALSE
myMatrixList <- mapply(.... , simplify=FALSE)
However, even that is not necessary, as you can just combine the sapply from the previous question and skip the middle step
I'd like to be able to subset my data.frame, DATA, into numeric fields and factor-type fields. My goal is to write generalized scripts to summarize my data, however, certain functions, such as hist or quantile, are not appropriate for non-numeric data. And it would make more sense to run table on the factor-type fields.
I tried using
types <- apply( DATA, 2, typeof)
to create a list of types for each field which I could then subset DATA by. However, this only caused errors. I'm sure there is a simply way of doing this but I've done a lot of searching and can't come up with anything.
Thanks.
[Since it worked, I'm posting my comment as an answer to this:]
Try lapply(DATA,class)