Numba嵌套列表类型定义问题求助

错误原因 #

你遇到的类型不兼容问题核心是两点：

1. permutations初始化为普通Python列表[]，但你往里面添加的是Numba typed List，两种类型无法统一；
2. Numba的typed嵌套列表需要明确声明内部元素类型，不能依赖自动推断。

方案1：使用Numba typed嵌套List #

通过预先创建类型模板，让Numba明确嵌套List的类型，同时统一内外层列表为typed List：

from numba.typed import List
from numba import njit
import numpy as np

@njit()
def build_permutations(position_length, starting_leverage, ending_leverage, starting_amount=100000):
    # 指定numpy数组类型，避免类型模糊
    current_permutation = np.full(position_length, starting_leverage, dtype=np.int64)
    
    # 创建子列表类型模板，告诉Numba内部是int64类型的List
    sublist_template = List()
    sublist_template.append(List(current_permutation))
    sublist_template.pop()
    
    # 初始化外层列表为嵌套List类型
    permutations = List.empty_list(sublist_template._type_)
    sublist = List()

    while True:
        total = (np.sum(current_permutation) * 440000) / starting_amount
        if 10 < total < 100:
            # 将numpy数组转为typed List并添加
            sublist.append(List(current_permutation))

            if len(sublist) == 100_000_000:
                permutations.append(sublist)
                sublist = List()

        # 生成下一个排列
        i = position_length - 1
        while i >= 0 and current_permutation[i] == ending_leverage:
            current_permutation[i] = starting_leverage
            i -= 1

        if i < 0:
            break

        current_permutation[i] += 1

    if sublist:
        permutations.append(sublist)

    return permutations

方案2：改用元组列表（更简洁） #

Numba对不可变元组的类型推断更友好，无需显式声明类型，代码更简洁：

from numba.typed import List
from numba import njit
import numpy as np

@njit()
def build_permutations(position_length, starting_leverage, ending_leverage, starting_amount=100000):
    current_permutation = np.full(position_length, starting_leverage, dtype=np.int64)
    
    # 初始化外层和内层为存元组的typed List
    permutation_type = tuple(current_permutation).__class__
    permutations = List.empty_list(permutation_type)
    sublist = List.empty_list(permutation_type)

    while True:
        total = (np.sum(current_permutation) * 440000) / starting_amount
        if 10 < total < 100:
            # 将numpy数组转为元组添加
            sublist.append(tuple(current_permutation))

            if len(sublist) == 100_000_000:
                permutations.append(sublist)
                sublist = List.empty_list(permutation_type)

        # 生成下一个排列
        i = position_length - 1
        while i >= 0 and current_permutation[i] == ending_leverage:
            current_permutation[i] = starting_leverage
            i -= 1

        if i < 0:
            break

        current_permutation[i] += 1

    if sublist:
        permutations.append(sublist)

    return permutations

额外提示 #

• 如果数据规模不大，也可以用普通Python列表存元组，但typed List在Numba JIT函数中的性能更优；
• 确保starting_leverage和ending_leverage是整数类型，避免numpy数组出现类型不一致问题。

内容的提问来源于stack exchange，提问作者sf8193