Databases
StackOverflow2013
Postsdbo
POnumpy FloatingPointError: invalid value encountered in subtract - not reproducible?
 

Note that there are some explanatory texts on larger screens.

plurals
  1. POnumpy FloatingPointError: invalid value encountered in subtract - not reproducible?
    primarykey
    data
    text
    <p>I'm encountering <code>FloatingPointError: invalid value encountered in subtract</code> in a piece of test code. The exception started being raised without any changes being made in the code itself, so I'm having a great deal of trouble understanding it.</p> <p>My question: What causes the <code>invalid value encountered in subtract</code> exception? Why would it behave differently on different installs of python+numpy?</p> <p>DETAILS:</p> <p>This MWE does <em>not</em> raise a <code>FloatingPointError</code>:</p> <pre><code>&gt;&gt;&gt; import numpy as np &gt;&gt;&gt; np.__version__ '1.6.1' &gt;&gt;&gt; x = np.arange(5,dtype='float64') &gt;&gt;&gt; y = np.ones(5,dtype='float64') &gt;&gt;&gt; x[2]=np.nan &gt;&gt;&gt; x-y # array([ -1., 0., nan, 2., 3.]) </code></pre> <p>However, deep within a piece of code, I subtract two <code>np.float64</code> <code>ndarray</code> objects, and get a floating point exception. The arrays causing the exception contain some pretty enormous and tiny numbers (e.g., 1e307 and 1e-307) and some <code>nan</code>s, but I haven't made any combination of these numbers result in an exception testing on my own.</p> <p>Much more disturbingly, I have a large grid of Jenkins tests running the exact same code with many versions of numpy, matplotlib, python, and scipy, and NONE of them raise this exception. I'm lost at this point - I don't know if there is a bug, or if there is, how to track it down. </p> <p>In case you're morbidly curious, the code in question is <a href="http://pyspeckit.bitbucket.org" rel="nofollow">pyspeckit</a> and the test is failing on line 20 of <code>test_hr2421.py</code>.</p> <p>EDIT: Follow-up - I think this little snippet: <code>np.seterr(invalid='raise')</code> was being called in a module I was importing, specifically pymc, and a pull request has since prevented this change from being made.</p>
    singulars
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. PTQuestion
    1. USkeflavich
    1. USkeflavich
    plurals
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. VO
      singulars
      1. POnumpy FloatingPointError: invalid value encountered in subtract - not reproducible?
      1. This table or related slice is empty.
      1. VTUpMod
    2. VO
      singulars
      1. POnumpy FloatingPointError: invalid value encountered in subtract - not reproducible?
      1. USRobert Muil
      1. VTFavorite
    3. VO
      singulars
      1. POnumpy FloatingPointError: invalid value encountered in subtract - not reproducible?
      1. This table or related slice is empty.
      1. VTApproveEditSuggestion
    1. COYou can raise the exception doing `np.seterr(invalid='raise');np.array(1e309) - np.array(1e309)` or ignore it setting `invalid='ignore'`. Thus, I'd say that somehow this is changing between versions.
      singulars
      1. POnumpy FloatingPointError: invalid value encountered in subtract - not reproducible?
      1. USjorgeca
    2. COThanks, that helped me track it a little further (or at least, pointed me in the right direction). Now the strangeness is that I'm taking `np.diff(np.array(arr))` (where `arr` is a subclass of `np.ndarray`) and the error (and nan values) occurs in the diff, but not in the input array.
      singulars
      1. POnumpy FloatingPointError: invalid value encountered in subtract - not reproducible?
      1. USkeflavich
    3. COThe array in np.diff, `a=asanyarray(a)` does not equal the input array; it contains NaNs and a variety of values that are tens to hundreds of orders of magnitude different. I found this out through the interactive debugger.
      singulars
      1. POnumpy FloatingPointError: invalid value encountered in subtract - not reproducible?
      1. USkeflavich
 

Querying!

 
Guidance
A row detail

Detail views are divided into sections. All the information in the data section comes from columns in the selected row. The other sections display data from other, related rows.

Related data can be related in a to-one or a to-many fashion. Captions of data related in a to-many fashion link to a list view showing a filtered view of the table.

Try moving around until you find a non-empty to-many entry and click on the label to get to one. You can move back to the root by clicking on the database name in the header.

SQuiL has stopped working due to an internal error.

If you are curious you may find further information in the browser console, which is accessible through the devtools (F12).

Reload