Add getindex; replace a.coeffs[i] with a[i] #97
Conversation
|
I just approved this to be merged, but it seems that tests (at least in 0.5) are not passing. I'll check this again. |
| return a[n+1]) | ||
|
|
||
| getindex(a::Taylor1, n::Int) = a.coeffs[n] | ||
| getindex(a::Taylor1, n::UnitRange) = a.coeffs[n] |
There was a problem hiding this comment.
Don't we need also methods for HomogeneousPolynomial and TaylorN? This may be the cause that the tests are not passing.
There was a problem hiding this comment.
Yes we do. I wanted to begin just with Taylor1 to get it right. I won't be able to get to the rest for a while if you want to have a go.
There was a problem hiding this comment.
I'll do it then.
One question: Should we also include setindex?
There was a problem hiding this comment.
Ah yes, I guess we should.
|
setindex! Is it surprising that the Taylor1 tests pass without setindex defined? |
|
I guess this is because we haven't done anything like |
|
But I changed all a.coeffs accesses to use getindex, so I'm not sure how
any entries in a TaylorSeries are getting changed...
…On 15 Mar 2017 10:26 a.m., "Luis Benet" ***@***.***> wrote:
I guess this is because we haven't done anything like a[1] = b[1]+c[1],
with all quantities Taylor1. This could simplify things even further!
Yet, Taylor1 and the rest are immutable, so I don't know if it works
completely...
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub
<#97 (comment)>,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/AALtTtOvrZrjAg8a0PSCpiEStJKCa7j3ks5rmBE0gaJpZM4Md_PQ>
.
|
|
Ah I see, they're immutable so you only use the constructor. Then there's
no need or use to define setindex!?
…On 15 Mar 2017 10:29 a.m., "David P. Sanders" ***@***.***> wrote:
But I changed all a.coeffs accesses to use getindex, so I'm not sure how
any entries in a TaylorSeries are getting changed...
On 15 Mar 2017 10:26 a.m., "Luis Benet" ***@***.***> wrote:
> I guess this is because we haven't done anything like a[1] = b[1]+c[1],
> with all quantities Taylor1. This could simplify things even further!
> Yet, Taylor1 and the rest are immutable, so I don't know if it works
> completely...
>
> —
> You are receiving this because you authored the thread.
> Reply to this email directly, view it on GitHub
> <#97 (comment)>,
> or mute the thread
> <https://github.com/notifications/unsubscribe-auth/AALtTtOvrZrjAg8a0PSCpiEStJKCa7j3ks5rmBE0gaJpZM4Md_PQ>
> .
>
|
|
Including |
|
I just pushed a new commit. With this functionality, I think we can simplify even further the code. I will work on that later (on a new PR)! I'll wait to get the green lights and then merge. |
|
Tests are passing on Travis for 0.5 and 0.6 (on Linux). |
dpsanders
changed the title
dpsanders
changed the title
|
I agree with you. One silly question: shall I rebase and merge, or squash and merge? |
|
I always just use the Squash and merge button. |
|
Thanks a lot!! |
a.coeffs[i]bya[i].getindexandsetindex!methods.