Skip to main content

LGPL library for small arms ballistic calculations (Python 3)

Project description

BallisticCalculator

LGPL library for small arms ballistic calculations based on point-mass (3 DoF) plus spin drift.

Table of contents

Installation

pip install py-ballisticcalc

# Using precompiled backend (improves performance)
pip install py-ballisticcalc[exts]

# Using matplotlib and pandas uses additional dependencies
pip install py-ballisticcalc[charts]

Usage

See Example.ipynb for detailed illustrations of all features and usage.

# Uncomment pyximport to compile instead of running pure python
#import pyximport; pyximport.install(language_level=3)

from py_ballisticcalc import *

Simple Zero

# Establish 100-yard zero for a standard .308, G7 bc=0.22, muzzle velocity 2600fps
zero = Shot(weapon=Weapon(sight_height=2), ammo=Ammo(DragModel(0.22, TableG7), mv=Velocity.FPS(2600)))
calc = Calculator()
zero_distance = Distance.Yard(100)
zero_elevation = calc.set_weapon_zero(zero, zero_distance)
print(f'Barrel elevation for {zero_distance} zero: {zero_elevation << PreferredUnits.adjustment}')
Barrel elevation for 100.0yd zero: 1.33mil

Plot Trajectory with Danger Space

# Plot trajectory out to 500 yards
shot_result = calc.fire(zero, trajectory_range=500, extra_data=True)
ax = shot_result.plot()
# Find danger space for a half-meter tall target at 300 yards
danger_space = shot_result.danger_space(Distance.Yard(300), Distance.Meter(.5))
print(danger_space)
danger_space.overlay(ax)  # Highlight danger space on the plot
plt.show()
Danger space at 300.0yd for 19.7inch tall target ranges from 217.1yd to 355.7yd

plot

Print Range Card

# Range card for this zero with 5mph cross-wind from left to right
zero.winds = [Wind(Velocity.MPH(5), Angular.OClock(3))]
range_card = calc.fire(zero, trajectory_range=1000)
range_card.dataframe().to_clipboard()
range_card.dataframe(True)[['distance', 'velocity', 'mach', 'time', 'target_drop', 'drop_adj', 'windage', 'windage_adj']].set_index('distance')
distance velocity mach time target_drop drop_adj windage windage_adj
0.0 yd 2600.0 ft/s 2.33 mach 0.000 s -2.0 inch 0.00 mil -0.0 inch 0.00 mil
100.0 yd 2398.1 ft/s 2.15 mach 0.120 s -0.0 inch -0.00 mil 0.4 inch 0.12 mil
200.0 yd 2205.5 ft/s 1.98 mach 0.251 s -4.1 inch -0.57 mil 1.7 inch 0.25 mil
300.0 yd 2022.3 ft/s 1.81 mach 0.393 s -15.3 inch -1.44 mil 4.1 inch 0.39 mil
400.0 yd 1847.5 ft/s 1.65 mach 0.548 s -35.0 inch -2.48 mil 7.6 inch 0.54 mil
500.0 yd 1680.1 ft/s 1.50 mach 0.718 s -65.0 inch -3.68 mil 12.4 inch 0.70 mil
600.0 yd 1519.5 ft/s 1.36 mach 0.906 s -107.3 inch -5.06 mil 18.8 inch 0.89 mil
700.0 yd 1366.0 ft/s 1.22 mach 1.114 s -164.8 inch -6.66 mil 27.0 inch 1.09 mil
800.0 yd 1221.3 ft/s 1.09 mach 1.347 s -240.9 inch -8.52 mil 37.3 inch 1.32 mil
900.0 yd 1093.2 ft/s 0.98 mach 1.607 s -340.5 inch -10.71 mil 50.0 inch 1.57 mil
1000.0 yd 1029.8 ft/s 0.92 mach 1.891 s -469.0 inch -13.27 mil 64.8 inch 1.83 mil

Complex Example

Here we define a standard .50BMG, enable powder temperature sensitivity, and zero for a distance of 500 meters, in a 5°C atmosphere at altitude 1000ft ASL.

dm = DragModel(0.62, TableG1, 661, 0.51, 2.3)
ammo=Ammo(dm, Velocity.MPS(850), Temperature.Celsius(15))
ammo.calc_powder_sens(Velocity.MPS(820), Temperature.Celsius(0))
weapon = Weapon(sight_height=Distance.Centimeter(9), twist=15)
atmo = Atmo(altitude=Distance.Foot(1000), temperature=Unit.Celsius(5), humidity=.5)
zero = Shot(weapon=weapon, ammo=ammo, atmo=atmo)
zero_distance = Distance.Meter(500)
calc = Calculator()
zero_elevation = calc.set_weapon_zero(zero, zero_distance)
print(f'Barrel elevation for {zero_distance} zero: {zero_elevation << PreferredUnits.adjustment}')
print(f'Muzzle velocity at zero temperature {atmo.temperature} is {ammo.get_velocity_for_temp(atmo.temperature) << Velocity.MPS}')
Barrel elevation for 500.0m zero: 4.69mil
Muzzle velocity at zero temperature 5.0°C is 830.0m/s

Units and Preferences

In version 2.x.x we changed concepts of settings, there are 2 ways to set preferences

1.1. To change library default units directly from code use PreferredUnits object

from py_ballisticcalc import PreferredUnits

# Change default library units
PreferredUnits.velocity = Velocity.MPS
PreferredUnits.adjustment = Angular.Mil
PreferredUnits.temperature = Temperature.Celsius
PreferredUnits.distance = Distance.Meter
PreferredUnits.sight_height = Distance.Centimeter
PreferredUnits.drop = Distance.Centimeter

print(f'PreferredUnits: {str(PreferredUnits)}')
print(f'Default distance unit: {PreferredUnits.distance.name}')

# Can create value in default unit with either float or another unit of same type
print(f'\tInstantiated from float (5): {PreferredUnits.distance(5)}')
print(f'\tInstantiated from Distance.Line(200): {PreferredUnits.distance(Distance.Line(200))}')

1.2. To change solver global setting use global flags setters

from py_ballisticcalc import *

# enable powder sensitivity calculation
set_global_use_powder_sensitivity(True)
# enable powder sensitivity calculation
set_global_max_calc_step_size(Unit.Meter(1))

# get that values
enabled = get_global_use_powder_sensitivity()
step = get_global_max_calc_step_size()

# reset global flags to defaults
reset_globals()

2. Use new method to set preferred units/settings globally for the venv or the user

Create .pybc.toml or pybc.toml file in your project root directory (where venv was placed). Or place this file in user's home directory. (The file in project root have priority.) You can use basicConfig() function to load your custom .toml file

The references of .pybc.toml settings file you can get there and there

# Config template for py_ballisticcalc

title = "standard py_ballisticcalc config template"
version = "2.0.0b4"

[pybc.preferred_units]
angular = 'Degree'
distance = 'Yard'
velocity = 'FPS'
# ... other there

[pybc.calculator]
max_calc_step_size = { value = 0.5, units = "Foot" }
use_powder_sensitivity = false
# ...
from py_ballisticcalc import basicConfig

basicConfig("path/to/your_config.toml")

Available manipulations with units

from py_ballisticcalc.unit import *

# Ways to define value in units
# 1. old syntax
unit_in_meter = Distance(100, Distance.Meter)
# 2. short syntax by Unit type class
unit_in_meter = Distance.Meter(100)
# 3. by Unit enum class
unit_in_meter = Unit.Meter(100)
print(f'100 meters: {unit_in_meter}')
# >>> 100 meters: 100.0m

# Convert unit
# 1. by .convert()
unit_in_yards = unit_in_meter.convert(Distance.Yard)
# 2. using shift syntax
unit_in_yards = unit_in_meter << Distance.Yard  # '<<=' operator also supports
print(f'100 meters in {unit_in_yards.units.key}: {unit_in_yards}')
# >>> 100 meters in yard: 109.4yd

# Get value in specified units (as float)
# 1. by .get_in()
value_in_km = unit_in_yards.get_in(Distance.Kilometer)
# 2. by shift syntax
value_in_km = unit_in_yards >> Distance.Kilometer  # '>>=' operator also supports
print(f'100 meters, value in km: {value_in_km}  (value type is {type(value_in_km)})')
# >>> 100 meters, value in km: 0.1  (value type is <class 'float'>)

# Getting unit raw value (a float)
rvalue = Distance.Meter(100).raw_value
rvalue = float(Distance.Meter(100))
print(f'100 meters in raw value: {rvalue}  (raw type is {type(rvalue)})')
# >>> 100 meters in raw value: 3937.0078740157483  (raw type is <class 'float'>)

# Comparison operators supported: < > <= >= == !=
print(f'Comparison: {unit_in_meter} == {Distance.Centimeter(100)}: {unit_in_meter == Distance.Centimeter(100)}')
# >>> False, compare two units by raw value
print(f'Comparison: {unit_in_meter} > .1*{unit_in_meter}: {unit_in_meter > .1*unit_in_meter.raw_value}')
# >>> True, compare unit with float by raw value

Concepts

Look angle

Look angle is the elevation of the sight line (a.k.a., Line of Sight, or LoS) relative to the horizon. For flat fire at angles close to horizontal this does not make a significant difference. When the look angle is significantly above or below the horizon the trajectory will be different because:

  1. Gravity is not orthogonal to the velocity
  2. Air density changes with altitude, so the drag effects will vary across an arcing trajectory.

The shooter typically cares about the line of sight (LoS): Sight adjustments (drop in the following figure) are made relative to LoS, and ranging errors – and hence danger space – follow the line of sight, not the horizon.

The following diagram shows how look distance and drop relate by look angle to the underlying (distance x, height y) trajectory data. Look-angle trigonometry

Danger Space

Danger space is a practical measure of sensitivity to ranging error. It is defined for a target of height h and distance d, and it indicates how far forward and backward along the line of sight the target can move such that the trajectory will still hit somewhere (vertically) on the target.

Danger Space

About project

The library provides trajectory calculation for ballistic projectiles including air rifles, bows, firearms, artillery, and so on.

The 3DoF model that is used in this calculator is rooted in public C code of JBM's calculator, ported to C#, optimized, fixed and extended with elements described in Litz's Applied Ballistics book and from the friendly project of Alexandre Trofimov and then ported to Go.

This Python3 implementation has been expanded to support multiple ballistic coefficients and custom drag functions, such as those derived from Doppler radar data.

The online version of Go documentation is located here.

C# version of the package is located here, and the online version of C# API documentation is located here.

Contributors

This project exists thanks to all the people who contribute.

Special thanks to:

  • David Bookstaber - Ballistics Expert
    For help understanding and improving the functionality
  • Nikolay Gekht
    For the sources code on C# and GO-lang from which this project firstly was forked

RISK NOTICE

The library performs very limited simulation of a complex physical process and so it performs a lot of approximations. Therefore, the calculation results MUST NOT be considered as completely and reliably reflecting actual behavior or characteristics of projectiles. While these results may be used for educational purpose, they must NOT be considered as reliable for the areas where incorrect calculation may cause making a wrong decision, financial harm, or can put a human life at risk.

THE CODE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE MATERIALS OR THE USE OR OTHER DEALINGS IN THE MATERIALS.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

py_ballisticcalc-2.0.4.tar.gz (257.5 kB view details)

Uploaded Source

Built Distribution

py_ballisticcalc-2.0.4-py3-none-any.whl (46.0 kB view details)

Uploaded Python 3

File details

Details for the file py_ballisticcalc-2.0.4.tar.gz.

File metadata

  • Download URL: py_ballisticcalc-2.0.4.tar.gz
  • Upload date:
  • Size: 257.5 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/5.1.1 CPython/3.12.4

File hashes

Hashes for py_ballisticcalc-2.0.4.tar.gz
Algorithm Hash digest
SHA256 9eaa5bf5f754065692af0b1be835384c7b16a0651570e9969029606b35d58ce1
MD5 7a3b3f3cf601e3cf196b2c849f378ec7
BLAKE2b-256 93f98ba1c9afc7e5a96c5ad9f1140386d3800a2acb1d6e3b5c142a058322cedc

See more details on using hashes here.

File details

Details for the file py_ballisticcalc-2.0.4-py3-none-any.whl.

File metadata

File hashes

Hashes for py_ballisticcalc-2.0.4-py3-none-any.whl
Algorithm Hash digest
SHA256 6033249732386bd2377834b594111587b554b1d4865fd45efdf16a591294ebbb
MD5 d32257f1f0a7ac8fced73d7903ed1824
BLAKE2b-256 290d2b8aa155b05164c3ef1bdec193b7a9d5a48af2bee4982c45b7eade42b63b

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page