Gottlob Frege is a multiversed man who has studied mathematics and philosophy. Perhaps it is not much of a surprise that he created a formal language,in which languages and mathematics found a common factor, known as logic. Frege's logic, as noted in his work Begriffsschrift was later improved on by men such as Bertrand Russel, Ludwig Wittgenstein and Willard van Orman Quine (among others).
In Frege's logic (and any other) it is clear that what is important are not the variables, but the connectives involved in any logical function. In reasoning appear precious few different connectives and these connectives are used to furmulate any logical judgement. An example of a logical judgement could be: "Whenever it rains I get wet", or any other thought what so ever. Frege names such thoughts "judgements".
"Judgements" are also called "predications" in logic. Some of these predications may in reality be contractions of several predications and the relations between them. An example might be that something is "all and one" at the same time, one would now use the word (predicate) "alone". Frege seperates predications in his work Über Funktion und Begriff. Frege seperates concepts from functions, where concepts are merely names where the predicate would be true or not depending if the object falls under the predication and functions are "truth values". Such "truth values" are based on the variables in the function. So the example "alone" would only be true if the thing one is "predicating" is "all and one". If one of the variable "predicates" would be untrue the outcome of the function would be untrue and the predicate alone would not fit the bill. The example shows one of the (very small) youthfull problems of Frege's invention called "logic". The example is one which can be both function and concept, depending on how one looks at things.
In his work "Über Sinn und Bedeutung" Frege defines two different aspects of "meaning". The first meaning of meaning is "reference"; what a thought-object refers to, and the second meaning is "sense"; how a thought-object relates to something. From this moment on one can clearly see that thought-objects are something else than the objects on which is thought. A clear stipulation of this is that one may, in thought, re-examine the thought-object and form a different opinion; reshaping the thought-object. In "reality" (whatever that may be) the object remains unchanged.
In his works "Über Funktion und Begriff" and "Über Sinn und Bedeutung" Frege sets out a theory with which one can understand how thought-objects are formed. When percieving any kind of object one "grasps" (Frege uses the word "fassen") some things out of something that one observed. This "grasping" refers to the process by which one forms a thought-object of an object. Grasping is involved in both reference (what object of all that one has percieved is one thinking of) and sense (in what way does it refer to the object).
From this point of view it is perhaps more clear that a predication-concept is used to form a thought-object by reference and a predication-function is used to form a thought-object by sense.

An illustration of "grasping". The magnified object is, in a way, "grasped" by the telescope (or the person wielding it).
What is also clear in Frege's work is that one "grasps" certain things from what one percieves. An important question might be what does this "grasping". In Frege's work the "grasping" takes the same place as 'urteilskraft' in Kant's work. Predications, or thought-objects, are, in this sense, the equivalent of Kant's Judgements.