Aap parent note ki ek bhi line padhne se pehle, aapko woh vocabulary chahiye jo woh aapse already pata hone ka assume karti hai. Yeh page har piece ko bilkul zero se build karta hai, us order mein jisme woh ek doosre par depend karte hain. Yahan koi bhi word use nahi hoga jiska pehle picture na draw ki ho.
Poora subject inhi do ke beech ke difference par jeeta hai. Position poochta hai "yeh line mein kaunsa wala hai?". Label poochta hai "iska naam kya hai?". Picture dekho: wahi teen boxes, do alag tareekon se address ki gayi hain.
Ek array mein koi name-tags nahi hote. Woh jaanta hai "2nd box mein 30 hai" lekin yeh nahi ki "u2 naam ke box mein 30 hai". Yahi missing name-tag woh gap hai jo pandas fill karta hai.
Ek missing box ko apni special value kyun chahiye, bas 0 kyun nahi? Kyunki 0 ek real answer hai ("humne zero sales measure ki"), jabki NaN ka matlab hai "humne kuch measure hi nahi kiya". Donon ko confuse karna har average ko corrupt kar deta hai jo aap compute karte hain. Parent ka alignment example dekho, jahan label "z" NaN produce karta hai.
Ab aap parent ka one-liner decode kar sakte hain: "aligned columns ki ek dictionary". Dictionary = column-name → column. Aligned = har column same row index share karta hai. Bas itna hi hai.
Upar se neeche padho: koi bhi cheez lower row par usable nahi hai jab tak usse feed karne wala har arrow samajh na aa jaye. Agar koi bhi upstream box fuzzy hai, parent note magic jaisa lagega.
Right side cover karo aur zor se jawab do. Agar ek bhi miss ho, uska section upar se dobara padho.
Value ki position aur uske label mein kya fark hai
Position = line mein kitna door hai (0,1,2…); label = name-tag jo chipka hua hai, jo rows reorder hone par kabhi nahi hilta.
Kaunse do arrays ek Series banate hain
Ek values array aur ek index (label) array jo equal length ke hain, ek saath stapled.
Iska kya matlab hai ki ek Series ek "partial function" hai
Yeh label → value map karta hai, lekin sirf unhi labels ke liye jo actually iske paas hain; missing labels ka koi answer nahi hota.
NaN kya hai aur sirf 0 kyun nahi use karte
Ek "Not a Number" marker jiska matlab hai yahan koi value nahi; 0 ek real measured value hai, NaN ka matlab hai kuch measure hi nahi kiya gaya.
Alignment rule plain words mein batao
Do Series ko label se line up karo; matching labels combine hote hain, unmatched labels NaN ban jaate hain.
Dictionary terms mein DataFrame kya hai
Ek dictionary {column label → Series} jahan har Series ek common row index share karti hai.
.loc slice inclusively kyun karta hai lekin .iloc exclusively
.loc labels use karta hai (ek naam ka koi "stop+1" nahi hota, isliye named endpoint include karo); .iloc positions use karta hai aur Python ke exclusive-stop rule ko follow karta hai.
Label lookup position lookup se almost utna hi fast kyun hai
Index ek hash table (naam→position) banata hai isliye naam-tag find karne ke liye poori strip scan nahi karni padti.
Kaunsa foundation directly groupby, merge, aur pivot ko power karta hai
Alignment — har ek "pehle label se match karo, phir act karo" hai.
