Metadata-Version: 2.1
Name: sjSDM-py
Version: 0.0.1
Summary: jSDM package
Home-page: https://github.com/TheoreticalEcology/s-jSDM/sjSDM/sjSDM-python
Author: Maximilian Pichler
Author-email: Maximilian.Pichler@ur.de
License: UNKNOWN
Platform: UNKNOWN
Classifier: License :: OSI Approved :: GNU General Public License v3 (GPLv3)
Classifier: Operating System :: OS Independent
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.5
Classifier: Programming Language :: Python :: 3.6
Classifier: Programming Language :: Python :: 3.7
Classifier: Programming Language :: Python :: 3.8
Description-Content-Type: text/markdown
Requires-Dist: numpy

# fajsm - Fast and Accurate Joint Species distribution Modeling 

## Install instructions
```{python}
pip install sjSDM_py
```

## Example
At the moment, we do not provide specifically a API for jSDM Models. However, it's just a deep multivariate probit model with one layer. Here's an example:

```{python}
import sjSDM_py as sa
import numpy as np
Env = np.random.randn(100, 5)
Occ = np.random.binomial(1, 0.5, [100, 10])

model = sa.Model_base(5) # input_shape == number of environmental predictors
model.add_layer(sa.layers.Layer_dense(hidden=10)) # number of hidden units in the layer == number of species
model.build(df=5, optimizer=sa.optimizer_adamax(lr=0.1, weight_decay = 0.01)) # df = degree of freedom 
model.fit(X = Env, Y = Occ)
print(model.weights_numpy)
print(model.get_cov())
```

* For species intercept, use 'bias=True' in 'Layer_dense(...)'. 
* We recommend to set df = number of species / 2. 

