pandas: Bin numbers with cut

Model pd.cut, which sorts numbers into labelled intervals, including its rules at the edges.

What to build

A program that bins values and counts each bin.

Requirements

  • Edges e0 < e1 < ... define intervals labelled in order. By default (right) a value x is in the interval (e_i, e_i+1]; with left it is in [e_i, e_i+1). With include_lowest (only meaningful with right), the first interval also holds e0.
  • A value in no interval is NaN. Print each value's label in input order, then LABEL: C for each label in order, then NaN: C.

Input

Line 1: the edges. Line 2: the labels (one fewer than the edges). Line 3: right, left or right include_lowest. Line 4: the values.

Output

The labels, then the counts.

Constraints

2 to 10 edges; 1 to 100 values; all numbers are whole numbers.

Example

Input:

0 17 64 120
child adult senior
right
5 18 30 64 70 0

Output:

child
adult
adult
adult
senior
NaN
child: 1
adult: 3
senior: 1
NaN: 1

Why it works

Intervals that include only one edge never overlap, so every value lands in at most one bin; the cost is that the outermost edge on the open side belongs to no bin unless include_lowest says otherwise.

Examples

Example 1 · Input
0 17 64 120
child adult senior
right
5 18 30 64 70 0
Expected output
child
adult
adult
adult
senior
NaN
child: 1
adult: 3
senior: 1
NaN: 1
Example 2 · Input
0 17 64 120
child adult senior
right include_lowest
0 17 120 121
Expected output
child
child
senior
NaN
child: 2
adult: 0
senior: 1
NaN: 1

Prompts inside input() are ignored when checking, so input("n: ") is fine.

Hints

Hint 1

With right=True an interval (a, b] holds a < x <= b; with right=False, [a, b) holds a <= x < b.

Hint 2

include_lowest=True also lets the very first interval hold its left edge.

Solution

Show one correct solution (try it yourself first)
edges = list(map(int, input().split()))
labels = input().split()
mode = input().split()
values = list(map(int, input().split()))
right = mode[0] == "right"
lowest = "include_lowest" in mode


def bin_of(x):
    for i in range(len(labels)):
        a, b = edges[i], edges[i + 1]
        if right and (a < x <= b or (lowest and i == 0 and x == a)):
            return labels[i]
        if not right and a <= x < b:
            return labels[i]
    return "NaN"


result = [bin_of(v) for v in values]
for r in result:
    print(r)
for label in labels:
    print(f"{label}: {result.count(label)}")
print(f"NaN: {result.count('NaN')}")

Lesson: Transforming columns →