WHY a limit? Because ∞ is not a number you can plug in. You integrate over a finite window [a,t] (perfectly legal) and then slide t→∞ to ask: does the accumulated area settle down?
WHY it works (first principles): Let F(t)=∫atf and G(t)=∫atg. Since f≥0, F is increasing. Since f≤g, F(t)≤G(t). A function that is increasing and bounded above must converge to a finite limit (Monotone Convergence). If ∫g converges, G(t) is bounded, so F(t) is bounded → ∫f converges. ∎
HOW to use it (the strategy):
Forecast: as x→∞, which power of x does f behave like?
Pick a clean g (usually x−p or e−x) on the correct side of the inequality.
Want to prove convergence? Bound fabove by a convergent g.
Want to prove divergence? Bound fbelow by a divergent g.
WHY: if f/g→L finite and positive, then for large x, 2Lg≤f≤2Lg. So f is squeezed between two constant multiples of g — DCT applies both ways, locking their fates together. HOW: just match the dominant terms; don't fight with inequalities.
Imagine pouring sand into a tall box that goes up forever. If a smaller pile of sand never overflows the floor (finite total), then a pile under it also stays finite. And if a pile bigger than yours overflows forever, then yours overflows too. The magic trick: instead of measuring your weird pile, you compare it to an easy pile (1/x2 stays finite, 1/x overflows) and just say "mine is smaller/bigger than that one!"
Dekho, improper integral matlab area infinity tak ja raha hai — sawaal yeh hai ki total area finite hai ya infinite. Seedha integrate karna kabhi-kabhi bahut tough hota hai, isliye hum comparison ka jugaad lagaate hain. Idea simple hai: agar tumhara function kisi aise function ke neeche baitha hai jiska area finite hai, to tumhara area bhi finite hoga. Aur agar tumhara function kisi aise function ke upar hai jiska area infinite hai, to tumhara bhi infinite ho jaayega.
Sabse important benchmark hai 1/xp. Yaad rakho: p>1 ho to converge, warna diverge. Direct Comparison Test (DCT) mein direction ka dhyaan rakhna — convergence prove karni hai to function ko ek convergent cheez ke neeche dabao (squeeze), aur divergence prove karni hai to ek divergent cheez ke upar dhakka do (push). Galat direction mein bound lagaya to kuch bhi conclude nahi hota — yeh sabse common galti hai.
Jab inequality banana mushkil ho (jaise x+x wali), tab Limit Comparison Test use karo: bas L=limf/g nikaalo, agar 0<L<∞ hai to dono ka fate same — dono converge ya dono diverge. Trick yeh hai ki g hamesha f ka dominant term pakdo (jaise large x pe x+x≈x, to g=1/x).
Yeh topic kyun important hai? Exams mein aksar poochte hain "converge karta hai ya nahi" bina actual value maange — wahan comparison seconds mein answer de deta hai. Aur yahi exact logic series ke comparison test mein bhi chalta hai, to ek baar samajh gaye to do chapters cover ho gaye. Mantra: "Squeeze to please, push to die."