My advice on this question: be careful, because there is a lot of misinformation.
At the end of the day, to "see" a color, you need light to come to your eye from some source. That source, in relation to your question, can be one of two things.
One: you shine broadband (white) light on a gas, the light of correct wavelengths gets absorbed by the atoms, and the rest passes through. You see the light that passes through, namely the "complementary" light. In practice, most of the colors get through, and you basically still see white light, unless you separate the colors somehow. Note that the emitted light from the atoms is in random directions and is typically much weaker than the input light, whereas the output complementary light is directional. If you don't look in the direction of the input beam, you'll only see the light emitted by the atoms when they decay, and then it looks kind of like the process described below.
Two: the atoms emit the light without an input. That is, something other than light excites them, and then in the process of decaying back to a lower energy state, they emit photons.
That is the source of color for halogens, neon signs, and most other things of definitely-colored light.
If you look up the colors for halogens, you'll be able to find the colors of their vapors "when heated." What this means is that the gas is heated, which cause its atoms to collide more often and more energetically. That energy is transferred into the electrons, which are excited. Then, when they naturally decay, they release the spectrum of colors you see.
Notice that no light went "in." Instead, the energy that excited the electrons begins as "thermal" energy. And the atom only emits one specific color, so you only see that one specific color. It is not absorption of light that is the source of color; it's heating and subsequent electron decay.