pandas: A model of DataFrame.info()

DataFrame.info() is the quickest overview of a table's structure. Model its report.

What to build

A program that prints an info-style report for a comma-separated table.

Requirements

  • Empty cells are missing. Print RangeIndex: N entries, 0 to N-1 (or RangeIndex: 0 entries for an empty table), then Data columns (total K columns):, then for each column I NAME C non-null DTYPE with I its position.
  • DTYPE uses these rules on the non-missing cells: bool if all are True/False (object if any are missing), int64 if all are whole numbers (float64 if any are missing), float64 if all are numbers, object otherwise; a column with no values is float64.
  • Finally print dtypes: TYPE(COUNT), ... for the dtypes used, in alphabetical order.

Input

Line 1: the header. Line 2: n. Then n rows, values separated by commas.

Output

The report.

Constraints

0 ≤ n ≤ 100; 1 ≤ columns ≤ 10.

Example

Input:

name,age,paid
3
Ada,36,True
Rahim,,False
,29,True

Output:

RangeIndex: 3 entries, 0 to 2
Data columns (total 3 columns):
0 name 2 non-null object
1 age 2 non-null float64
2 paid 3 non-null bool
dtypes: bool(1), float64(1), object(1)

Why it works

info combines the shape, the missing-value counts and the dtypes, the three facts that decide what cleaning a table needs, which is why it is usually the first thing printed in an analysis.

Examples

Example 1 · Input
name,age,paid
3
Ada,36,True
Rahim,,False
,29,True
Expected output
RangeIndex: 3 entries, 0 to 2
Data columns (total 3 columns):
0 name 2 non-null object
1 age 2 non-null float64
2 paid 3 non-null bool
dtypes: bool(1), float64(1), object(1)
Example 2 · Input
x,y
2
1,2.5
3,4
Expected output
RangeIndex: 2 entries, 0 to 1
Data columns (total 2 columns):
0 x 2 non-null int64
1 y 2 non-null float64
dtypes: float64(1), int64(1)

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

Hints

Hint 1

Count the non-empty cells of each column; empty cells are missing.

Hint 2

Reuse the dtype rules: bool, int64 (float64 with missing values), float64, otherwise object.

Solution

Show one correct solution (try it yourself first)
def is_number(text, kind):
    try:
        kind(text)
        return True
    except ValueError:
        return False


def dtype_of(cells):
    present = [c for c in cells if c != ""]
    missing = len(cells) - len(present)
    if not present:
        return "float64"
    if all(c in ("True", "False") for c in present):
        return "object" if missing else "bool"
    if all(is_number(c, int) for c in present):
        return "float64" if missing else "int64"
    if all(is_number(c, float) for c in present):
        return "float64"
    return "object"


header = input().split(",")
n = int(input())
rows = [input().split(",") for _ in range(n)]
print(f"RangeIndex: {n} entries, 0 to {n - 1}" if n else "RangeIndex: 0 entries")
print(f"Data columns (total {len(header)} columns):")
used = {}
for i, name in enumerate(header):
    cells = [row[i] for row in rows]
    dtype = dtype_of(cells)
    used[dtype] = used.get(dtype, 0) + 1
    print(f"{i} {name} {sum(1 for c in cells if c != '')} non-null {dtype}")
print("dtypes: " + ", ".join(f"{t}({used[t]})" for t in sorted(used)))

Lesson: DataFrames: tables of labelled columns →