Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/ryu577/statest
Statistical estimation
https://github.com/ryu577/statest
Last synced: 7 days ago
JSON representation
Statistical estimation
- Host: GitHub
- URL: https://github.com/ryu577/statest
- Owner: ryu577
- License: mit
- Created: 2022-01-03T08:26:51.000Z (almost 3 years ago)
- Default Branch: main
- Last Pushed: 2022-01-10T05:37:43.000Z (almost 3 years ago)
- Last Synced: 2024-09-22T13:17:02.382Z (about 2 months ago)
- Language: Python
- Size: 37.1 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# statest
A library for statistical estimation of the properties of various distributions.
## Installation
> pip install statest
## First use-case: estimate quantiles
The R quantile function has 9 estimators. The Python estimator in numpy (percentile) has only one (the seventh version in R). This library implements the remaining 8 methods.
```python
from statest.quantile.estimate import est_1, est_2, est_3,\
est_4, est_5, est_6,\
est_7, est_8, est_9
a = np.arange(15); q=0.8
quartile = est_4(a, q)
```
To measure the performance of the quantile estimation methods on various distributions, the library has a simulator. The simulations take about 20 minutes to run. The bias, variance and MSE are plotted for all the estimators. The data from the simulations is stored to disk in a folder called "sim_data" and the plots are saved to a folder called "plots".
```python
from statest.quantile.simulator.estimator_racetrack1 import RaceTrack
rt = RaceTrack("Normal")
rt.race()
```
The avaiable distributions are: Normal, LogNormal, LogLogistic, Weibull, Lomax and Exponential.
## Resources
Based on the paper: https://arxiv.org/abs/2201.01421 and the blog: https://towardsdatascience.com/hear-me-out-i-found-a-better-way-to-estimate-the-median-5c4971be4278