我有一个csv文件:
Index,X1,X2,X3,X4,X5,Y
1,-1.608052,-0.377992,1.204209,1.313808,1.218265,1
2,0.393766,0.630685,-1.222062,0.090558,0.015893,0
3,-0.466243,0.276972,2.519047,0.673745,0.16729,1
4,1.47121,-0.046791,-0.303291,-0.365437,1.989287,0
5,-1.672906,1.25588,-0.355706,0.123143,-2.241941,1
我想创建一个分类系统程序,数据在第二行。我想从第二排得到数据。我试着用下一个(列表)这样:
def load_DataTrain(filename):
try:
with open(filename, newline='') as iFile:
return list(reader(iFile, delimiter=','))
next(list)
except FileNotFoundError as e:
raise e
但它不工作,我得到一个错误,因为程序从第一行读取。我没有使用pandas或csv.reader来读取我的csv。这是我从Divyesh GitHub得到的代码:
from csv import reader
from sys import exit
from math import sqrt
from operator import itemgetter
def load_DataTrain(filename):
try:
with open(filename) as iFile:
return list(reader(iFile, delimiter=','))
next(list)
except FileNotFoundError as e:
raise e
def convert_to_float(DataTrain, mode):
new_set = []
try:
if mode == 'training':
for data in DataTrain:
new_set.append([float(x) for x in data[:len(data)-1]] + [data[len(data)-1]])
elif mode == 'test':
for data in DataTrain:
new_set.append([float(x) for x in data])
else:
print('Invalid mode, program will exit.')
exit()
return new_set
except ValueError as v:
print(v)
print('Invalid data set format, program will exit.')
exit()
def get_classes(training_set):
return list(set([c[-1] for c in training_set]))
def find_neighbors(distances, k):
return distances[0:k]
def find_response(neighbors, classes):
votes = [0] * len(classes)
for instance in neighbors:
for ctr, c in enumerate(classes):
if instance[-2] == c:
votes[ctr] += 1
return max(enumerate(votes), key=itemgetter(1))
def knn(training_set, test_set, k):
distances = []
dist = 0
limit = len(training_set[0]) - 1
# generate response classes from training data
classes = get_classes(training_set)
try:
for test_instance in test_set:
for row in training_set:
for x, y in zip(row[:limit], test_instance):
dist += (x-y) * (x-y)
distances.append(row + [sqrt(dist)])
dist = 0
distances.sort(key=itemgetter(len(distances[0])-1))
# find k nearest neighbors
neighbors = find_neighbors(distances, k)
# get the class with maximum votes
index, value = find_response(neighbors, classes)
# Display prediction
print('The predicted class for sample ' + str(test_instance) + ' is : ' + classes[index])
print('Number of votes : ' + str(value) + ' out of ' + str(k))
# empty the distance list
distances.clear()
except Exception as e:
print(e)
def main():
try:
# get value of k
k = int(input('Enter the value of k : '))
# load the training and test data set
training_file = input('Enter name of training data file : ')
test_file = input('Enter name of test data file : ')
training_set = convert_to_float(load_DataTrain(training_file), 'training')
test_set = convert_to_float(load_DataTrain(test_file), 'test')
if not training_set:
print('Empty training set')
elif not test_set:
print('Empty test set')
elif k > len(training_set):
print('Expected number of neighbors is higher than number of training data instances')
else:
knn(training_set, test_set, k)
except ValueError as v:
print(v)
except FileNotFoundError:
print('File not found')
if __name__ == '__main__':
main()
结果是:
could not convert string to float: 'Index'
我应该怎么做才能从csv文件的第二行读取?
4条答案
按热度按时间i2loujxw1#
你的功能发生了微小的变化。
如果你只想返回
2nd
行,那么你可以在下面的代码中将[1:]
替换为[1]
。输出:
使用
pandas
的替代方案将
df.values.tolist()
更改为df.iloc[0].values.tolist()
,以仅返回2nd
行。输出:
kxxlusnw2#
s5a0g9ez3#
既然你在你的例子中使用了iris数据集,我猜你是在冒险进入机器学习?如果是这样的话,我认为使用pandas来读取和处理.csv文件会更明智。
4uqofj5v4#
使用基本Python(无导入)