tuple to numpy, data accuracy

Question

When I convert a tuple to numpy, there is a problem with data accuracy. My code is like this:

import numpy as np
a=(0.547693688614422, -0.7854270889025808, 0.6267478456110592)
print(a)
print(type(a))
tmp=np.array(a)
print(tmp)

The result is like this:

(0.547693688614422, -0.7854270889025808, 0.6267478456110592)
<class 'tuple'>
[ 0.54769369 -0.78542709  0.62674785]

Why?

Asked By: June-Solstice

||

Source

Answer 1

One way is to set this:

In [1039]: np.set_printoptions(precision=20)

In [1041]: tmp=np.array(a)

In [1042]: tmp
Out[1042]: array([ 0.547693688614422 , -0.7854270889025808,  0.6267478456110592])

In [1043]: tmp.dtype
Out[1043]: dtype('float64')

Answered By: Mayank Porwal

Answer 2

I think you’re only seeing a truncation in display only, but the internal value still retains the original accuracy. Here’s what I found:

>> a
(0.547693688614422, -0.7854270889025808, 0.6267478456110592)

>> b=np.array(a)

>> b
array([ 0.54769369, -0.78542709,  0.62674785]) #<-- printed display shows lower accuracy

>> b[0]
0.547693688614422 #<-- print of a single value shows same accuracy as original

So there’s no reason to change any settings – math performed with these arrays will still be at full accuracy.

Answered By: Demis

Answer 3

This seeming discrepancy should just be how the numbers are being displayed, not how they are being represented / stored.

You can check the dtype to verify it is still float64

tmp.dtype  # dtype('float64')

You can adjust np.set_printoptions to see them values displayed differently

print(tmp)  # [ 0.54769369 -0.78542709  0.62674785]
np.set_printoptions(precision=18)  # default precision is 8
print(tmp)  # [ 0.547693688614422  -0.7854270889025808  0.6267478456110592]

Answered By: Capybara

tuple to numpy, data accuracy

Question:

Answers: