NumPy: np.histogram by hand

np.histogram splits the data's range into equal bins and counts the values in each. Reproduce its counts and edges.

What to build

A program that prints histogram counts and bin edges.

Requirements

  • Print counts followed by the counts in NumPy's 1-D format.
  • Then print edges followed by the bins + 1 edges, each in Python's :g format, separated by spaces.
  • Bins include their left edge; the last bin also includes the maximum. When all values are equal, use the range min − 0.5 to max + 0.5.

Input

Line 1: the data. Line 2: the number of bins.

Output

Two lines.

Constraints

1 to 1000 values between −10⁶ and 10⁶; 1 ≤ bins ≤ 50.

Example

Input:

1 2 2 3 4
3

Output:

counts [1 2 2]
edges 1 2 3 4

Why it works

Equal-width bins make a value's bin a simple division, and closing the last bin keeps the maximum inside the range instead of creating an extra bin for it.

Examples

Example 1 · Input
1 2 2 3 4
3
Expected output
counts [1 2 2]
edges 1 2 3 4
Example 2 · Input
0 10
2
Expected output
counts [1 1]
edges 0 5 10

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

Hints

Hint 1

width = (max − min) ÷ bins, and a value's bin is int((x − min) ÷ width).

Hint 2

Put the maximum into the last bin by taking the smaller of that index and bins − 1.

Hint 3

If all values are equal, NumPy uses the range min − 0.5 to max + 0.5.

Solution

Show one correct solution (try it yourself first)
data = [float(x) for x in input().split()]
bins = int(input())
lo, hi = min(data), max(data)
if lo == hi:
    lo, hi = lo - 0.5, hi + 0.5
width = (hi - lo) / bins
counts = [0] * bins
for x in data:
    counts[min(int((x - lo) / width), bins - 1)] += 1
w = max(len(str(c)) for c in counts)
print("counts [" + " ".join(str(c).rjust(w) for c in counts) + "]")
print("edges", " ".join(f"{lo + i * width:g}" for i in range(bins + 1)))

Lesson: Statistics with NumPy →