thresh=n，保留至少有 n 个非 NA 数的行
df
0 1 2
0 0.035669 NaN NaN
1 -2.138661 NaN NaN
2 1.500880 NaN -0.687359
3 1.573421 NaN 1.041288
4 -0.305428 -0.482399 0.215099
5 -0.572742 1.043271 -0.492441
6 -0.222994 -0.781335 0.762350df.dropna(thresh=1)
0 1 2
0 0.035669 NaN NaN
1 -2.138661 NaN NaN
2 1.500880 NaN -0.687359
3 1.573421 NaN 1.041288
4 -0.305428 -0.482399 0.215099
5 -0.572742 1.043271 -0.492441
6 -0.222994 -0.781335 0.762350df.dropna(thresh=3)
0 1 2
4 -0.305428 -0.482399 0.215099
5 -0.572742 1.043271 -0.492441
6 -0.222994 -0.781335 0.762350df.dropna(thresh=2)
0 1 2
2 1.500880 NaN -0.687359
3 1.573421 NaN 1.041288
4 -0.305428 -0.482399 0.215099
5 -0.572742 1.043271 -0.492441
6 -0.222994 -0.781335 0.762350df.dropna(thresh=5)
Empty DataFrame
Columns: [0, 1, 2]
Index: []