先放个示范代码:
# coding:utf-8
import yaml
import networkx as nx
import numpy as np
import scipy.io
file_path = r"\graph\WS100_4_0.1.yaml"
with open(file_path, "r") as fd:
G = yaml.load(fd)
A = np.matrix(nx.adjacency_matrix(G).todense(), dtype=bool)
print(type(A), type(A[0,0]))
scipy.io.savemat('WS.mat', mdict={'A':A})
--------------------------------------------------
python中导出矩阵至matlab
一、通过scipy.io直接存为mat(推荐)
例如
import scipy.io
scipy.io.savemat('filename', mdict={ 'a':a, 'b':b})
在matlab中只需要
load 'filename';
就导入了a矩阵和b矩阵
python存储矩阵
import pandas as pd
df = pd.DataFrame(a)
df.to_csv("score",sep=" ",index = False)
二、通过文本文件
如果矩阵是mxn维的。 那么可以用 :
np.savetxt('dev_ivector.csv', dev_ivector, delimiter = ',')
对应matlab读取为:
dev_ivec = csvread('dev_ivector.csv') ###csv格式其实就内定了结构体
如果矩阵是(n,)这种格式。['aagj' 'aagy' 'aann' ... 'zzgm' 'zzhk' 'zzwn'] 类似这种。那么可以用
f = open('label','w')
for x in spk_mean_label:
print(x)
print(x,file=f)
f.close()
对应matlab读取为:
spk_mean_label = importdata('label')
-----------------------------------------------------
matlab中导出矩阵至python
一、通过 scipy.io 读取mat文件
matlab里面得到矩阵后可以直接从工作区变量处保存为.mat文件。
然后在python中执行下面步骤
import scipy.io as sio
load_fn = 'plda_bl_score.mat'
load_data = sio.loadmat(load_fn)
blscores= load_data['PLDA_bl_scores']
然后blscores就是矩阵了。
假如有俩个矩阵score,score1
save('score.mat','score','score1')
在python中
import scipy.io
matlab_data = scipy.io.loadmat('score.mat')
score = matlab_data['score']
score1 = matlab_data['score1']
转自:https://blog.csdn.net/fukaiqi1010/article/details/81939566