[collection of knowledge points] contour coefficient, Unicode to float, reading CSV file skip the first line

Profile factor

Intuitive understanding ： It's probably a measure of whether samples of the same category can be gathered together （ That is to visualize the clustering results , Can you clearly see that there is an obvious aggregation of categories with the same color ）

disadvantages ： Sometimes the contour coefficient is high , There may be a risk of clustering too fine （ Such as apple 、 The contour coefficient of pears is not as good as that of apples alone 、 The contour coefficient of pear cluster is high , The latter is usually what we don't want to see ）

python Lieutenant general unicode Turn into float

Read python2 Generated .npy When you file , Often read unicode, Processing mode ：

v= float(cur.encode("utf-8"))

python Read .csv File skips the first line

Reference article Python csv Skip the first line

import csv
from itertools import islice
with open(' form /2019-04-01.csv', 'r') as read_file:
    reader = csv.reader(read_file)
    for row in islice(reader, 1, None):
        print(row)

I said why I read it before .csv When , Saved after reading .npy or .json Often one more ”id“