0
# Regime Models
1

2
Spatial regression models allowing parameters to vary across regimes, with options for separate or joint estimation and extensive regime-specific diagnostics.
3

4
## Capabilities
5

6
### Spatial Error Regime Models
7

8
```python { .api }
9
class GM_Error_Het_Regimes:
10
    def __init__(self, y, x, regimes, w, constant_regi='many', cols2regi='all',
11
                 regime_err_sep=False, regime_lag_sep=False, cores=False,
12
                 max_iter=1, epsilon=0.0000001, step1c=False, 
13
                 inv_method='power_exp', hard_bound=False, vm=False, 
14
                 name_y=None, name_x=None, name_regimes=None, name_w=None,
15
                 name_ds=None, latex=False):
16
        """
17
        GMM spatial error model with regimes and heteroskedasticity.
18

19
        Parameters:
20
        - y (array): nx1 dependent variable
21
        - x (array): nxk independent variables
22
        - regimes (list/Series): Regime identifier for each observation
23
        - w (sparse matrix): Spatial weights matrix
24
        - constant_regi (str): 'one' (same constant) or 'many' (different per regime)
25
        - cols2regi (str/list): Variables that vary by regime ('all' or list)
26
        - regime_err_sep (bool): Separate error variance by regime
27
        - regime_lag_sep (bool): Separate spatial parameter by regime
28
        - cores (bool): Use multiprocessing for regime estimation
29

30
        Attributes:
31
        - regimes_set (set): Unique regime values
32
        - kr (int): Number of regimized variables  
33
        - nr (int): Number of regimes
34
        - multi (dict): Individual regime results (if regime_err_sep=True)
35
        - chow (dict): Chow test for regime differences
36
        """
37

38
class GM_Error_Hom_Regimes:
39
    def __init__(self, y, x, regimes, w, constant_regi='many', cols2regi='all',
40
                 regime_err_sep=False, regime_lag_sep=False, cores=False,
41
                 hard_bound=False, vm=False, name_y=None, name_x=None,
42
                 name_regimes=None, name_w=None, name_ds=None, latex=False):
43
        """GMM spatial error model with regimes assuming homoskedasticity."""
44
```
45

46
## Usage Examples
47

48
### Basic Regime Model
49

50
```python
51
import numpy as np
52
import spreg
53
from libpysal import weights
54

55
# Data with regimes
56
n = 150
57
x = np.random.randn(n, 2)
58
y = np.random.randn(n, 1)
59
regimes = np.random.choice(['North', 'South', 'East'], n)
60
w = weights.KNN.from_array(np.random.randn(n, 2), k=5)
61

62
# Spatial error model with regimes
63
regime_model = spreg.GM_Error_Het_Regimes(y, x, regimes, w.sparse,
64
                                          constant_regi='many', cols2regi='all',
65
                                          name_y='y', name_x=['x1', 'x2'],
66
                                          name_regimes='region')
67

68
print(regime_model.summary)
69
print(f"Number of regimes: {regime_model.nr}")
70
print("Chow test for regime differences:", regime_model.chow)
71
```