Python 목록에서 ndarray까지

NumPy's core gift is the ndarray, a grid of numbers all of one type. It is the workhorse behind almost every data tool you'll use. 🚀

A Python list can hold anything: numbers, strings, even other lists. That freedom is handy, but for crunching numbers it makes the list slow.

An array trades flexibility for speed. Every element shares one dtype, so NumPy stores them tightly and computes on them at near-C speed.

You build an array by passing a list to np.array. NumPy reads the values and wraps them into a fast numeric grid for you.

Mix an int and a float in one array and NumPy quietly promotes everything to float, because all elements must share a single type.

Because the type is fixed, NumPy skips the per-item checks a list needs. This vectorization is why arrays fly through millions of numbers. ⚡

List items scatter across memory, but an array sits in one contiguous block. Your CPU loves that, reading values in tight, predictable bursts.

You don't always start from a list. np.zeros and np.ones fill an array with a chosen value, perfect for placeholders you fill in later.

Need an even sequence? np.arange works like Python's range but hands you back an array ready for math.

An array isn't limited to one row. Pass nested lists and you get a 2D grid, the matrix shape that powers tables and images.

pandas, scikit-learn, and matplotlib all sit on top of the ndarray. Master this one type and the whole data stack opens up. 🧱

Let's check the key difference between lists and arrays.

You met the ndarray: a single-type, tightly packed grid that beats lists for numbers. Build it with np.array, zeros, or arange, then let the speed work for you. 🎉