[Freeciv-Dev] Re: [PATCH] [1.2] move some common AI equations to kill_desire() (PR#1279)

[Raahul Kumar <raahul_da_man@xxxxxxxxx>]

--- Petr Baudis <pasky@xxxxxxxxxxx> wrote:
> Dear diary, on Mon, Feb 25, 2002 at 01:41:39PM CET, I got a letter,
> where Raahul Kumar <raahul_da_man@xxxxxxxxx> told me, that...
> > > --- advmilitary.c 22 Feb 2002 19:32:40 -0000      1.1.1.9
> > > +++ advmilitary.c 24 Feb 2002 21:36:20 -0000      1.2
> > > @@ -620,7 +620,7 @@
> > >                unit_types[i].move_type == HELI_MOVING) && acity &&
> > >                acity->ai.invasion == 2) b0 = f * SHIELD_WEIGHTING;
> > >        else {
> > > -        b0 = (b * a - (f + (acity ? acity->ai.f : 0)) * d) * g *
> > > SHIELD_WEIGHTING / (a + g * d);
> > > +        b0 = kill_desire(b, a, (f + (acity ? acity->ai.f : 0)), d, g);
> > >          if (acity && b * acity->ai.a * acity->ai.a > acity->ai.f * d)
> > >            b0 -= (b * acity->ai.a * acity->ai.a - acity->ai.f * d) *
> > >                             g * SHIELD_WEIGHTING / (acity->ai.a *
> acity->ai.a
> > > + g * d);
> > 
> > Kill those a,b,c,d. I think you have better names for those.
> 
> I do (basically it looks that same substitution as in findvictim applies) -
> however, as I said, it won't be in this patch.

Sounds good. I'm curious about the patches you intend to send in to Freeciv
dev. How are you going to split them up ? I want to be ahead of Raimar with
insightful comments. Never easy, he seems to be reading the patches and
responding first.

> 
> > > --- aiunit.c      23 Feb 2002 21:03:31 -0000      1.1.1.10
> > > +++ aiunit.c      24 Feb 2002 21:36:20 -0000      1.45
> > > @@ -616,6 +616,30 @@
> > >  }
> > >  
> > > 
> /**************************************************************************
> > > +Compute how much we want to kill certain victim we've chosen.
> > > +
> > > +Benefit is something like 'attractiveness' of the victim, how nice it
> would
> > > be
> > > +to destroy it. Larger value, worse loss for enemy.
> > > +
> > > +Attack is the total possible attack power we can throw on the victim.
> Note
> > > that
> > > +we will even square this.
> > > +
> > > +Loss is the possible loss when we would lose the unit we want to attack.
> > > +
> > > +Vuln is vulnerability of our unit when attacking the enemy.
> > >
> +**************************************************************************/
> > 
> > Victim count is the only one missing. Victim count does not count stacked
> units
> > right as far as I know.
> 
> Please elaborate.
> 
> I'll add description of victim count, thanks.
> 
> Improved patch attached (only this changes).
>

You seem to know more about victim count than I do. If you have a specific
question, ask.
 



>  /**************************************************************************
> +Compute how much we want to kill certain victim we've chosen.
> +
> +Benefit is something like 'attractiveness' of the victim, how nice it would
> be
> +to destroy it. Larger value, worse loss for enemy.
> +
> +Attack is the total possible attack power we can throw on the victim. Note
> that
> +we will even square this.
> +
> +Loss is the possible loss when we would lose the unit we want to attack.
> +
> +Vuln is vulnerability of our unit when attacking the enemy.
> +
> +Victim count is number of victims stacked in the target tile. FIXME?: The
> +equation is not accurate as the other values can vary for other units on the
> +target tile (we take values from best defender), however I believe it's
> +accurate just enough now and lost speed isn't worth that.

I can't tell from this comment what happens when a fortress is the target tile.

> +**************************************************************************/
> +int kill_desire(int benefit, int attack, int loss, int vuln, int
> victim_count)
> +{
> +  int desire;
> +
> +  /*         attractiveness     danger */ 
> +  desire = ((benefit * attack - loss * vuln) * victim_count *
> SHIELD_WEIGHTING
> +            / (attack + vuln * victim_count));
> +
> +  return desire;
> +}
> +

Excellent.

> +/**************************************************************************
>  Military "want" estimates are amortized in this complicated way.
>  COMMENTME: Why not use simple amortize? -- GB
>  **************************************************************************/
> @@ -805,20 +834,10 @@
>                  unit_owner(pdef)->name, unit_type(pdef)->name);
>          
>        } else {
> +        /* See description of kill_desire() about this variables. */
>          int vuln = unit_vulnerability(punit, pdef);
> -        
> -        /* The total possible attack power we can throw on the victim. Note
> -         * that we will even square this. */
>          int attack = reinforcements_value(punit, pdef->x, pdef->y) + bellig;
> -        
> -        /* Something like 'attractiveness' of the victim, how nice it would
> be
> -         * to destroy it. Larger value, worse loss for enemy. */
>          int benefit = unit_type(pdef)->build_cost;
> -        
> -        /* We're only dealing with adjacent victims here. */
> -        int move_cost = 0;
> -        
> -        /* The possible loss when we would lose the unit we want to attack.
> */
>          int loss = unit_type(punit)->build_cost;
>          
>          attack *= attack;
> @@ -836,15 +855,10 @@
>          /* If we have non-zero belligerence... */
>          if (attack > 0 && is_my_turn(punit, pdef)) {
>            int desire;
> -          
> -          /* TODO: This equation is simplified version of much worse ones in
> -           * that long fat routines, but it's still a common pattern, so we
> -           * will can this equation to one separate readable function. We'll
> -           * also be able to remove move_cost and loss variables. */
> -         
> -          /*         attractiveness     danger */ 
> -          desire = ((benefit * attack - loss * vuln) * SHIELD_WEIGHTING
> -                    / (attack + vuln) - move_cost * SHIELD_WEIGHTING);
> +
> +          /* FIXME? Why we don't use stack_size as victim_count? --pasky */
> +
> +          desire = kill_desire(benefit, attack, loss, vuln, 1);
>            
>            /* No need to amortize! We're doing it in one-turn horizon (?). */
>            
> @@ -1520,7 +1534,7 @@
>            else if ((is_ground_unit(punit) || is_heli_unit(punit)) &&
>                     acity->ai.invasion == 2) b0 = f * SHIELD_WEIGHTING;
>            else {
> -            b0 = (b * a - (f + acity->ai.f) * d) * g * SHIELD_WEIGHTING / (a
> + g * d);
> +            b0 = kill_desire(b, a, (f + acity->ai.f), d, g);
>              if (b * acity->ai.a * acity->ai.a > acity->ai.f * d)
>                 b0 -= (b * acity->ai.a * acity->ai.a - acity->ai.f * d) *
>                        g * SHIELD_WEIGHTING / (acity->ai.a * acity->ai.a + g
> * d);
> Index: aiunit.h
> ===================================================================
> RCS file: /home/cvs/aiciv/freeciv-a2/ai/aiunit.h,v
> retrieving revision 1.1.1.4
> retrieving revision 1.7
> diff -u -r1.1.1.4 -r1.7
> --- aiunit.h  19 Feb 2002 19:33:54 -0000      1.1.1.4
> +++ aiunit.h  24 Feb 2002 21:36:21 -0000      1.7
> @@ -37,6 +37,7 @@
>  int unit_vulnerability_basic(struct unit *punit, struct unit *pdef);
>  int unit_vulnerability_virtual(struct unit *punit);
>  int unit_vulnerability(struct unit *punit, struct unit *pdef);
> +int kill_desire(int benefit, int attack, int loss, int vuln, int
> attack_count);
>  int military_amortize(int value, int delay, int build_cost);
>  
>  bool is_on_unit_upgrade_path(Unit_Type_id test, Unit_Type_id base);
> 


Seems good to me. I vote for inclusion, Tony, Greg, Ross?


__________________________________________________
Do You Yahoo!?
Yahoo! Sports - Coverage of the 2002 Olympic Games
http://sports.yahoo.com