How do you arrive at choke losses in street light fittings what is total wattage what is ballast wattage?

Assuming the question is about simple reactive control gear, e.g. a choke or inductor in series with the discharge lamp, then the following answer applies.
In an ideal situation, an ideal inductor, that is an inductor exhibiting pure inductance by virtue of its construction and the use of wire with no resistance, then the losses would be zero and the power consumed by the circuit would be V.I.Cos Phi, where Phi is the phase angle between the voltage and current.
In the real world electronic components are rarely ever perfect, and although a component may be called an inductor or choke, it will possess some parasitic resistance and possibly some stray capacitance too. The main source of the parasitic resistance is due to the resistance of the wire used to wind it, the diameter and length of the wire required to obtain a particular inductance and current rating and any eddy-current losses in an iron core if one is used. In the case of street lighting control gear designed to function at 50 or 60Hz, laminated iron cores are almost always used to obtain the relatively large inductances required. (At 240v R.M.S., 50Hz approximately 1 Henry of inductance is required to regulate a 20 Watt discharge lamp and 100mH for 200Watt lamp and 10mH for a 2Kw lamp, particular lamp types such as fluorescent, high pressure sodium or metal halide of the same power rating requiring slightly differing inductance values dependent on their discharge voltages at their specified operating currents.) The laminated iron cores are often constructed with an air gap to control the current value at which the core saturates.
In a real choke power is wasted in I squared R losses in the parasitic resistance of the wire, which appears as a low value, (hundreds of ohms for low current, high inductance chokes down to a an ohm or less for high current low inductance chokes used for large H.I.D. lamps) in series with the actual reactive inductance obtained by the number of turns, thickness of the wire and the nature of the core. Eddy current losses appear as a parallel resistance across the series combination of the actual inductance and resistance of the wire. Both of these resistive components add a small in-phase current to the large lagging current of the inductive component. (An ideal inductor exhibits a pure quadrature lagging current.) The series resistive component due to the wire is easy to measure and can simply be measured with a multi-meter on the appropriate resistance range for relatively high inductance chokes used with lamps up to about 400w rating. A four wire milli-ohm meter may be needed to measure the lower resistances of "larger" , lower inductance chokes with fewer turns of thicker wire, used with lamps rated above 400w.
Simply measuring the operating current, squaring it and multiplying this by the previously measured resistance will give the loss in the choke due to its wire resistance, but this will not include any eddy current losses in the core. Another method would be to connect the choke directly across a low source impedance A.C. voltage of the intended operating frequency. Generally most series regulating discharge lamp chokes can be connected directly across the supply voltage they are intended to be used from, (lamp shorted out), although some may saturate and in this case the voltage would need to be reduced. Under these conditions one could use a low value shunt or a suitably terminated current transformer in series with the choke under test to obtain the current waveform on an oscilloscope. The voltage waveform across the choke could also be displayed on the second channel of the oscilloscope to give both the amplitudes and phase angle between the voltage an current. Rembering to take into account the ratios of the current and voltage transformers used and to convert the peak values to R.M.S., the in-phase component should be able to be extracted by resolving the phase ange and magnitudes from the oscilloscope into two orthogonal components. The magnitude of this in-phase component of choke current can be used to calculate all the lossy resistive components, Rt, of a real choke and thus using "Iop" squared "Rt" to get the trus losses. ("Iop", being the operating current through the choke when used with a fully run-up lamp in series.)
Some notes, low inductance, high current chokes wound on large cores with few turns of thick wire and used for H.I.D. lamps with power ratings of aroud 1Kw or more are a lot less lossy than high inductance, low current chokes used for small fluorescent tubes which are wound with many thousands of turns of fine wire with high resistance. High frequency control gear, such as that used in electronic compact fluorescent lamps. operating at frequencies of around 45-50KHz make use of the fact that inductive reactence = 2pifL, and by increasing the frequency, f, the inductance, L, can be reduced by an inverse amount and since the inductance, L, is roughly proportional to the number of turns, these can be reduced requiring a lot less wire and thus possessing a lot less resistance and thus are a lot less lossy. The losses in the rectifier and oscillator used to raise the frequency are more than compensated for by the large increase in inductance.
Measuring omic "I squared R" losses in control gear such as leak transformers and "constant wattage", CWA gear is more demanding but not impossible.