Data for feature importance calculation

tree_data.py

# Entries in the nodes 
n_entries = {
    "node 1": 15480,
    "node 2": 12163,
    "node 3": 3317,
    "node 4": 5869,
    "node 5": 6294,
    "node 6": 2317,
    "node 7": 1000, 
    "node 8": 2454,
    "node 9": 3415,
    "node 10": 1372,
    "node 11": 4922,
    "node 12": 958,
    "node 13": 1359,
    "node 14": 423,
    "node 15": 577
}

# Nodes squared errors
i_sq = {
    "node 1": 1.335,
    "node 2": 0.832,
    "node 3": 1.214,
    "node 4": 0.546,
    "node 5": 0.834,
    "node 6": 0.893,
    "node 7": 0.776, 
    "node 8": 0.648,
    "node 9": 0.385,
    "node 10": 1.287,
    "node 11": 0.516,
    "node 12": 0.989,
    "node 13": 0.536,
    "node 14": 0.773,
    "node 15": 0.432
}

# Defining the features used in the nodes that have a splitting rule
feature_in_node = {
    "node 1": "MedInc",
    "node 2": "MedInc",
    "node 3": "MedInc",
    "node 4": "AveRooms",
    "node 5": "AveOccup",
    "node 6": "AveOccup",
    "node 7": "MedInc"
}

# The first node's count is the total number of entries in the tree
_n = n_entries["node 1"]

# Calculating the n entries probability
n_entries_weight = {}
for node in n_entries:
    n_entries_weight[node] = n_entries[node] / _n

# Defining the relationship between nodes; 
# The key is the parent node, and the value is a list of 
# [left child node, right child node]
node_pairs = {
    "node 1": ["node 2", "node 3"],
    "node 2": ["node 4", "node 5"],
    "node 3": ["node 6", "node 7"],
    "node 4": ["node 8", "node 9"],
    "node 5": ["node 10", "node 11"],
    "node 6": ["node 12", "node 13"],
    "node 7": ["node 14", "node 15"]
}