Re: [eigen] Re: smallest()

To: eigen <eigen@xxxxxxxxxxxxxxxxxxx>
Subject: Re: [eigen] Re: smallest()
From: Gael Guennebaud <gael.guennebaud@xxxxxxxxx>
Date: Thu, 15 Sep 2016 11:08:36 +0200
Dkim-signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20120113; h=mime-version:in-reply-to:references:from:date:message-id:subject:to; bh=z7+tFlwER1Q3gGYHp7bELGQqgZAtn+zF/SZfDunAoVw=; b=CSivDiIJnSsG2d+74m+AmAy6ae3LGgXDeLAbpYM5fMPpUPRiVtP4ylGv90WaG/e4dP A872EBXYUvx10wNOB/e/wmzBTBnukqef8N1IEJhv+aOYS8FGBr6suM9VfuI09y40H9Yi zlga2NG5rzE84YfnYoLly57tejzFbCYK7U+MzceIOBi2HvLkuvFTxvNNoF27loQjnWHD z9rufH8zB+oIpQMk++6OOS9c9f3oZfN6eiI7DtwsSgZJ3LZHls6H9ljako4mpDuyUeFI 6lFMcCO2FcVbq80xp/HX6oiN5DxlqjO+Np4hOS19qOgOBUanybxvPWItpA8D84B8s7Ui agBQ==

On Mon, Sep 12, 2016 at 6:34 PM, Peter <list@xxxxxxxxxxxxxxxxx> wrote:

Documenting it is probably a good idea, as forgetting to specialize numeric_limits produces wrong code for custom types without any warning.

I'll add a static assertion on numeric_limits<T>::is_specialized.

BTW, will one also need the const and the reference versions, like numeric_limits<const TpRFloat> , or are they never called?

no no, only one is needed!!!

gael.

References:
- [eigen] Householder.h: ::min() and operator<=
  - From: Peter
- Re: [eigen] Householder.h: ::min() and operator<=
  - From: Christoph Hertzberg
- smalleset() ( Re: [eigen] Householder.h: ::min() and operator<=)
  - From: Peter
- Re: smalleset() ( Re: [eigen] Householder.h: ::min() and operator<=)
  - From: Christoph Hertzberg
- Re: smalleset() ( Re: [eigen] Householder.h: ::min() and operator<=)
  - From: Peter
- Re: smalleset() ( Re: [eigen] Householder.h: ::min() and operator<=)
  - From: Gael Guennebaud
- Re: smalleset() ( Re: [eigen] Householder.h: ::min() and operator<=)
  - From: Pavel Holoborodko
- Re: smalleset() ( Re: [eigen] Householder.h: ::min() and operator<=)
  - From: Christoph Hertzberg
- [eigen] Re: smallest()
  - From: Peter

Mail converted by MHonArc 2.6.19+