Mercurial > pylearn

comparison doc/v2_planning/architecture.txt @ 1153:ae5ba6206fd3

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
a first draft of pseudo-code for logreg .. using version B (?) approach
author Razvan Pascanu <r.pascanu@gmail.com>
date Thu, 16 Sep 2010 17:34:30 -0400
parents 40c1461ce9ef
children f111f8c2a280
comparison
equal deleted inserted replaced
1152:0904dd74894d 1153:ae5ba6206fd3
114 (*) Note that I cheated a bit in my example above: the graph from the 114 (*) Note that I cheated a bit in my example above: the graph from the
115 encapsulated version is not a simple chain, so it is not obvious how to 115 encapsulated version is not a simple chain, so it is not obvious how to
116 convert it into the pipeline given in the linear version. It's still possible 116 convert it into the pipeline given in the linear version. It's still possible
117 though, but this is probably not the place to get into the details. 117 though, but this is probably not the place to get into the details.
118 118
119 RP comment : The way I see it, you could always have everything using the
120 encapsulation paradigm ( which as you pointed out is a bit more powerful) and
121 then have linear shortcuts ( functions that take a list of functions and some
122 inputs and apply them in some order). You will not be able to have a one case
123 cover all pipeline function, but I think it is sufficient to offer such
124 options (linear functions) for a few widely used cases ..
119 125
120 126
121 Jobman Compatibility Approach 127 Jobman Compatibility Approach
122 ============================= 128 =============================
123 129