pyneatR 0.1.1

Neat formatting for numbers, dates, and strings using numpy.

pyneatR

Neat formatting for numbers, dates, timestamps, and strings using numpy. Implementation inspired by the R package neatR.

Installation

pip install pyneatR

Neat Data for Presentation

Formatting dates

Often, we encounter dates in various formats and want an unambiguous representation. ndate formats dates into a readable mmm dd, yyyy format with the day of the week.

from pyneatR import ndate, nday
from datetime import date, timedelta

today = date.today()
# Assuming today is Nov 09, 2025 (Sun) for these examples

print(ndate(today - timedelta(days=3)))
# "Nov 06, 2025 (Thu)"

print(ndate(today))
# "Nov 09, 2025 (Sun)"

print(ndate(today + timedelta(days=4)))
# "Nov 13, 2025 (Thu)"

To just get the date without the day of week, set display_weekday=False:

print(ndate(today, display_weekday=False))
# "Nov 09, 2025"

For monthly data, abbreviating the date to mmm'yy is elegant:

print(ndate(today, display_weekday=False, is_month=True))
# "Nov'25"

To see the context of the date with respect to current date (within 1 week), use nday.

print(nday(today, reference_alias=False))
# "Sun"

print(nday(today, reference_alias=True))
# "Today, Sun"

Formatting timestamp

Timestamps are feature-rich representations of date and time.

from pyneatR import ntimestamp
from datetime import datetime

now = datetime.now()
# Assuming now is Nov 09, 2025 12:07:48 PM

print(ntimestamp(now))
# "Nov 09, 2025 12H 07M 48S PM (Sun)"

To extract and format only the time:

print(ntimestamp(now, display_weekday=False, include_date=False, include_timezone=False))
# "12H 07M 48S PM"

Components of time can be toggled on or off:

print(ntimestamp(now, include_date=False, display_weekday=False, 
                 include_hours=True, include_minutes=True, include_seconds=False, 
                 include_timezone=False))
# "12H 07M PM"

Formatting number

nnumber formats numeric data in a readable way, automatically handling large numbers.

from pyneatR import nnumber
import numpy as np

x = np.array([10, 100, 1000, 10000, 100000, 1000000, 10000000, 100000000, 1000000000])

print(nnumber(x))
# ['10.0' '100.0' '1.0 K' '10.0 K' '100.0 K' '1.0 Mn' '10.0 Mn' '100.0 Mn' '1.0 Bn']

print(nnumber(x, digits=0))
# ['10' '100' '1 K' '10 K' '100 K' '1 Mn' '10 Mn' '100 Mn' '1 Bn']

nnumber can automatically determine the best single unit for all numbers using unit='auto':

x = np.array([1e6, 99e3, 76e3, 42e3, 12e3, 789, 53])
print(nnumber(x, unit='auto'))
# ['1,000 K' '99 K' '76 K' '42 K' '12 K' '0.8 K' '0.1 K']

You can specify the unit directly:

print(nnumber(123456789.123456, unit='Mn'))
# "123.5 Mn"

Default units are 'K', 'Mn', 'Bn', 'Tn'. Custom labels can be provided:

print(nnumber(123456789.123456, unit='M', unit_labels={'million': 'M'}))
# "123.5 M"

Prefixes and suffixes can be added:

print(nnumber(123456789.123456, unit='Mn', prefix='$ '))
# "$ 123.5 Mn"

To show the raw number with separators:

print(nnumber(123456789.123456, digits=2, unit='', thousand_separator=','))
# "123,456,789.12"

Formatting percentages

npercent formats percentages, handling both decimals (0.228) and values already multiplied by 100 (22.8).

from pyneatR import npercent

print(npercent(22.8, is_decimal=False))
# "+22.8%"

print(npercent(0.228, is_decimal=True))
# "+22.8%"

By default is_decimal=True. The plus sign can be toggled:

print(npercent(0.228, plus_sign=False))
# "22.8%"

For growth factors:

tesla_2017 = 20
tesla_2023 = 200
g = (tesla_2023 - tesla_2017) / tesla_2017

print(npercent(g, plus_sign=True))
# "+900.0%"

# factor_out=True is not yet implemented in this version, but check documentation for updates.

Formatting string

nstring formats character vectors, removing special characters or changing case.

from pyneatR import nstring

s = ' All MOdels are wrong. some ARE useful!!! '
print(nstring(s, case='title', remove_specials=True))
# "All Models Are Wrong Some Are Useful"

To retain only English alphabets and numbers:

print(nstring(s, case='title', remove_specials=True, en_only=True))
# "All Models Are Wrong Some Are Useful"