记录一些面试的题吧，没事自己看看
1.python,c++,java棋盘中的白子进行八连通区域查找，并返回连通区域块数和连通区域分类情况
# encoding=utf8
def show(L):
# 直观显示用 非本题核心代码
rows = len(L)
cols = len(L[0])
for i in range(rows):
for j in range(cols):
print(L[i][j], end=" ")
print()
# 棋盘19*19 1为白色棋子 0为其他
L = [[0, 0, 0, 0, 1, 1, 1, 0, 0, 0, 0, 0, 1, 1, 1, 0, 1, 0, 1],
[0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 1, 1, 0, 1, 0, 1],
[0, 1, 1, 1, 1, 1, 1, 0, 0, 1, 1, 0, 1, 1, 1, 0, 1, 0, 1],
[0, 0, 1, 0, 1, 0, 0, 0, 0, 1, 1, 0, 1, 1, 1, 0, 1, 0, 1],
[0, 0, 0, 0, 0, 1, 1, 0, 0, 1, 1, 0, 1, 1, 1, 0, 1, 0, 1],
[0, 0, 0, 0, 1, 1, 1, 0, 0, 1, 1, 0, 1, 1, 1, 0, 1, 0, 1],
[0, 1, 1, 0, 1, 1, 1, 0, 0, 1, 1, 0, 1, 1, 1, 0, 1, 0, 1],
[0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 1, 1, 0, 1, 0, 1],
[0, 1, 1, 0, 1, 1, 1, 0, 0, 1, 1, 0, 1, 1, 1, 0, 1, 0, 1],
[0, 0, 1, 0, 1, 0, 0, 0, 0, 1, 1, 0, 1, 1, 1, 0, 1, 0, 1],
[0, 0, 0, 0, 0, 1, 1, 0, 0, 1, 1, 0, 1, 1, 1, 0, 1, 0, 1],
[0, 0, 0, 0, 1, 1, 1, 0, 0, 1, 1, 0, 1, 1, 1, 0, 1, 0, 1],
[0, 0, 0, 0, 1, 1, 1, 0, 0, 1, 1, 0, 1, 1, 1, 0, 1, 0, 1],
[0, 0, 0, 0, 1, 1, 1, 0, 0, 1, 1, 0, 1, 1, 1, 0, 1, 0, 1],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 1, 0, 1],
[0, 0, 0, 0, 1, 1, 1, 0, 0, 1, 1, 0, 0, 1, 0, 0, 0, 0, 1],
[0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 0, 1, 1, 1, 0, 1, 1, 1],
[0, 0, 0, 0, 1, 1, 1, 0, 0, 1, 1, 0, 1, 1, 1, 0, 1, 0, 1],
[0, 0, 0, 0, 1, 1, 1, 0, 0, 1, 1, 0, 1, 1, 1, 0, 1, 0, 1]]
rows = len(L)
cols = len(L[0])
# 生成flag矩阵 标记处理过的点
flag = []
for i in range(rows):
col = cols * [0]
flag.append(col)
def means(i, j, k):
'''
聚类查找
:param i: 棋盘行坐标
:param j: 棋盘列坐标
:param k: 聚类类别
'''
# 非白棋或者查找过的点排除
if i < 0 or i >= rows or j < 0 or j >= cols or \
flag[i][j] != 0 or L[i][j] != 1:
return 0
flag[i][j] = k
# 使用八连通区域查找
means(i, j - 1, k)
means(i, j + 1, k)
means(i - 1, j, k)
means(i + 1, j, k)
means(i + 1, j - 1, k)
means(i + 1, j + 1, k)
means(i - 1, j - 1, k)
means(i - 1, j + 1, k)
if __name__ == "__main__":
# k聚类
k = 1
for i in range(rows):
for j in range(cols):
# 未查找过的白棋
if L[i][j] == 1 and flag[i][j] == 0:
means(i, j, k)
k += 1
print('共有{}块'.format(k))
print('聚类效果如下：')
show(flag)
##########完整的展示#############
#!/usr/bin/python
# encoding=utf8
table = [[0, 0, 1, 0, 1, 1, 1, 0],
[0, 0, 1, 0, 0, 1, 1, 0],
[0, 1, 1, 0, 1, 1, 1, 0],
[0, 0, 1, 0, 1, 0, 0, 0],
[0, 0, 0, 0, 0, 1, 1, 0],
[0, 0, 0, 0, 1, 1, 1, 0]]
rows = len(table)
cols = len(table[0])
label_table = []
for i in range(rows):
col = cols * [0]
label_table.append(col)
def show(table):
rows = len(table)
cols = len(table[0])
for i in range(rows):
for j in range(cols):
print(table[i][j], end=" ")
print()
def dfs(i, j, mask):
if i < 0 or i >= rows or j < 0 or j >= cols or \
label_table[i][j] != 0 or \
table[i][j] != 1:
return 0
label_table[i][j] = mask
ret = 1
# left right up down search
ret += dfs(i, j - 1, mask)
ret += dfs(i, j + 1, mask)
ret += dfs(i - 1, j, mask)
ret += dfs(i + 1, j, mask)
return ret
if __name__ == "__main__":
print("original table:")
show(table)
res = {}
print("++++++++++++++++++++")
print("label table")
mask = 1
for i in range(rows):
for j in range(cols):
if table[i][j] == 1 and label_table[i][j] == 0:
ret = dfs(i, j, mask)
res[mask] = ret
mask += 1
show(label_table)
print("++++++++++++++++++++")
print("results:")
sorted_res = [(k, res[k]) for k in sorted(res, key=res.get, reverse=True)]
max_grp = sorted_res[0][0]
print("max group num: %d" % sorted_res[0][1])
for i in range(rows):
for j in range(cols):
if label_table[i][j] == max_grp:
print("point (%d, %d) belongs to max group: %d" % (i, j, max_grp))
2.python二维矩阵转置，不能使用numpy
# encoding=utf8
def show(L):
# 直观显示用 非本题核心代码
rows = len(L)
cols = len(L[0])
for i in range(rows):
for j in range(cols):
print(L[i][j], end=" ")
print()
def transpose(A):
B = []
for i in range(len(A[0])):
col = [0] * len(A)
B.append(col)
for i in range(len(A)):
for j in range(len(A[0])):
B[j][i] = A[i][j]
print('转置前：')
show(A)
print('转置后：')
show(B)
if __name__ == '__main__':
A = [[1, 2, 3],
[4, 5, 6]]
transpose(A)
3.python计算两个bbox的iou 人家没说不能用numpy我就用了
# encoding=utf8
import numpy as np
def bbox_iou(gt_bbox, pred_bbox):
'''
:param gt_bbox: 数据集bbox坐标
:param pred_bbox: 预测bbox坐标
:return: iou是否符合阈值 阈值默认0.5
'''
# 得到union区域左上角和右下角
insert_left = np.maximum(gt_bbox[:2], pred_bbox[:2])
insert_right = np.minimum(gt_bbox[2:], pred_bbox[2:])
# 重合区域面积
insert_area = (insert_right[0] - insert_left[0]) * (insert_right[1] - insert_left[1])
# gt_bbox面积
gt_area = (gt_bbox[2] - gt_bbox[0]) * (gt_bbox[3] - gt_bbox[1])
# pred_bbox面积
pred_area = (pred_bbox[2] - pred_bbox[0]) * (pred_bbox[3] - pred_bbox[1])
iou = insert_area / (gt_area + pred_area - insert_area)
if iou > 0.5:
return True
else:
return False
if __name__ == '__main__':
# 例子
gt_bbox = [0.1, 0.1, 0.5, 0.5]
pred_bbox = [0.3, 0.2, 0.7, 0.9]
if bbox_iou(gt_bbox, pred_bbox):
print('预测框达到IOU阈值')
else:
print('预测框未达到IOU阈值')
4.top_k查找 不允许使用sorted和list.sort
David 89
Tony 66
Tom 56
Jack 98
William 78
# encoding=utf8
def find_good_gay(file_name):
'''
scores.txt文件在本项目文件夹中
:param file_name: 文件路径
:return: 成绩最好的三个学生的姓名和成绩
'''
# 读取txt文件
f = open(file_name)
L = f.readlines()
num = len(L)
# 使用字典格式处理数据
res_dict = {}
# 生成成绩字典
for i in range(num):
res_dict[L[i].split(' ')[0]] = L[i].strip().split(' ')[1]
# 排序
for i in range(num):
for j in range(num - i - 1):
key1 = list(res_dict.keys())[j]
key2 = list(res_dict.keys())[j + 1]
if res_dict[key1] < res_dict[key2]:
res_dict[key1], res_dict[key2] = res_dict[key2], res_dict[key1]
print('成绩排在前三名的分别是：')
for i in range(3):
print(list(res_dict.keys())[i] + ':' + res_dict[list(res_dict.keys())[i]])
if __name__ == '__main__':
find_good_gay('scores.txt')
5.python+tnsorflow 完成线性回归的搭建（可以用tf写伪代码），这个代码没法运行也算是个伪代码吧
import tensorflow as tf
x_train = tf.constant([1, 2, 3, 4])
y_train = tf.constant([2, 4, 6, 8])
with tf.device('/gpu:0'):
x = tf.placeholder(dtype=tf.float32, shape=[4,], name='x')
y = tf.plcceholder(dtype=tf.float32, shape=[4,], name='y')
w = tf.get_variable(name='w', shape=[4,], dtype=tf.float32, initializer=tf.random_normal_initializer(stddev=0.1, mean=0))
b = tf.get_variable(name='b', shape=[4,], dtype=tf.float32, initializer=tf.zeros_initializer())
y_ = tf.matmul(x, w) + b
loss = tf.reduce_mean(tf.square(y_ - y))
# loss1 = tf.nn.sparse_softmax_cross_entropy_with_logits(labels=y, logits=y_)
optimizer = tf.train.GradientDescentOptimizer(0.001).minimize(loss)
saver = tf.train.Saver()
sess = tf.Session(tf.ConfigProto(allow_soft_placement=True))
init = tf.global_variables_initializer()
sess.run(init)
for i in range(100):
_, loss = sess.run([optimizer, loss], feed_dict={x: x_train, y: y_train})
print('Train Step Loss:', loss)
saver.save(sess,save_path='path')
浙公网安备 33010602011771号