字典numpy arrays到（转）csv

Question

I've a dictionary of numpy float arrays:

labels = 'L0', 'L1', 'L2'
x = {'a': array([1.0, 1.1, 1.2]), 'b':array([2.0, 2.1, 2.2])}

I'd like to get a.csv file something like

'Label','a', 'b'
'L1',1.0, 2.0
'L2',1.1, 2.1
'L3',1.2, 2.2

New to Python... I'm somewhat lost... Thx. for your help!

Answer 1

Pandas has exactly the thing. You will love it

import pandas as pd

dataframe = pd.DataFrame(x, index=labels)
dataframe.to_csv('filename.csv')

Answer 2

You can use pythons csv-module for that. You would need to specify quotechar="'" and quoting=csv.QUOTE_NONNUMERIC to ensure the quoting you wanted:

import numpy as np

labels = 'L0', 'L1', 'L2'
x = {'a': np.array([1.0, 1.1, 1.2]), 'b':np.array([2.0, 2.1, 2.2])}

import csv

with open("d.txt","w", newline="") as f:
    # ensure quoting of non numerics
    wr = csv.writer(f, quotechar="'", quoting=csv.QUOTE_NONNUMERIC)
    header = ["Label"] + [k for k in x.keys()]
    wr.writerow(header)
    data = zip(labels,*x.values())
    for d in data:
        wr.writerow(d)

print(open("d.txt").read())

Output of file:

'Label','a','b'
'L0',1.0,2.0
'L1',1.1,2.1
'L2',1.2,2.2