graphlab创建sframe如何获取SArray中位数

Question

I'm studying graphlab create with

data=graphlab.SFrame.read_csv('test.csv')

im trying to get median of one of columns

data_train.fillna(('Credit_History',data_train['Credit_History'].median()))

but I got error

---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
<ipython-input-247-50ed3eb09dcc> in <module>()
----> 1 data_train.fillna(('Credit_History',data_train['Credit_History'].median()))

AttributeError: 'SArray' object has no attribute 'median'

data.show() will show median of this column though anyone knows how to fix this?

Answer 1

I think I understand what your trying to do. Sframe doesn't have a default median function. I would improvise like this:

import numpy as np
data_train.fillna('Credit_History', np.median(data_train['Credit_History']))

Answer 2

SArray doesn't have a median method. The best way to get the median is through the sketch_summary method, then quantile . More info on the sketch summary at

https://turi.com/products/create/docs/generated/graphlab.Sketch.html

import numpy as np
import graphlab as gl

sf = gl.SFrame(np.random.rand(100))

sketch = sf['X1'].sketch_summary()
median = sketch.quantile(0.5)