Is it more efficient to use del or dictionary comprehension in python?

Question

Suppose I have a python dictionary with the following structure:

data = {
"key_1": "some value",
"key_2": "some value",
"key_3": "",
"key_4": ""
}

So, I want to remove all key-values for which the value is an empty string. I could use the python one in the following way:

for key in list(data.keys()):
    if data[key] == "":
        del data[key]

Or I could build a new dictionary with a comprehension:

new_ dict = {
            key: value
            for key, value in data.items()
            if value != ""
        }

Both solutions are very readable, but which is more efficient and why?