https://github.com/arose13/rosey-sim

little package for very basic sensitivity analysis
https://github.com/arose13/rosey-sim

Last synced: 12 days ago
JSON representation

little package for very basic sensitivity analysis

• Host: GitHub
• URL: https://github.com/arose13/rosey-sim
• Owner: arose13
• Created: 2020-06-17T01:52:43.000Z (over 4 years ago)
• Default Branch: master
• Last Pushed: 2020-06-17T14:58:10.000Z (over 4 years ago)
• Last Synced: 2024-11-29T15:54:02.922Z (29 days ago)
• Language: Python
• Size: 4.88 KB
• Stars: 0
• Watchers: 3
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# Rosey-sim

Causal, Probabilistic and Uncertainty Simulations

## Installation

```bash

pip install rosey-sim

```

## Example

```python

from rosey_sim.sensitivity import Model, Parameter

from tqdm import trange

import matplotlib.pyplot as graph

import seaborn as sns

# Specify function

def profit_func(price, cost):

    return price - cost

# Specify model (NOTE the parameters matches the functions arguments)

profit_model = Model(

    profit_func,

    params=[

        Parameter('cost', [3, 7]),

        Parameter('price', [10, 12])

    ]

)

# Single sample

print(f'${profit_model.sample():.2f}')

# Many single samples

trace = [profit_model.sample() for _ in trange(10000)]

sns.distplot(trace, label='One at a time')

# Many sample immediately

trace = profit_model.sample(10000)

sns.distplot(trace, label='All at once')

graph.show()

# Even nesting is allowed

def random_divide(profit_func, denom):

    return profit_func / denom

final_model = Model(

    random_divide,

    params=[

        profit_model,

        Parameter('denom', [3, 4])

    ]

)

print(final_model.sample())

```