Skip to content

Commit 9f3dedc

Browse files
committed
Add compute pages
1 parent 97e34a6 commit 9f3dedc

File tree

4 files changed

+58
-0
lines changed

4 files changed

+58
-0
lines changed

docs/src/compute/distributed.md

Lines changed: 3 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,3 @@
1+
## Work in progress
2+
You may ask questions in the chat window below or
3+
refer to [legacy documentation](https://docs.datajoint.org/)

docs/src/compute/key-source.md

Lines changed: 3 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,3 @@
1+
## Work in progress
2+
You may ask questions in the chat window below or
3+
refer to [legacy documentation](https://docs.datajoint.org/)

docs/src/compute/make.md

Lines changed: 49 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,49 @@
1+
# Make Method
2+
3+
For auto-populated *Imported* and *Computed* tables[^1], a `make` method gives exact
4+
instructions for generating the content. By making these steps explicit, we keep a
5+
careful record of data provenance and ensure reproducibility. Data should never be
6+
entered using the `insert` method directly.
7+
8+
[^1]: For information on differentiating these data tiers, see the Table Tier section on
9+
[Automation](../tabletiers#automation-imported-and-computed).
10+
11+
The `make` method receives one argument: the *key*, which represents the upstream table
12+
entries that need populating. The `key` is a `dict` or `struct` in Python and Matlab,
13+
respectively.
14+
15+
A `make` function should do three things:
16+
17+
1. [Fetch](../../query-lang/common-commands#fetch) data from tables upstream in the
18+
pipeline using the key for restriction.
19+
20+
2. Compute and add any missing attributes to the fields already in the key.
21+
22+
3. [Inserts](../../query-lang/common-commands#insert) the entire entity into the
23+
triggering table.
24+
25+
## Populate
26+
27+
The `make` method is sometimes referred to as the `populate` function because this is
28+
the class method called to run the `make` method on all relevant keys[^2].
29+
30+
[^2]: For information on reprocessing keys that resulted in an error, see information
31+
on the [Jobs table](../../ref-integrity/distributed-computing).
32+
33+
=== "Python"
34+
35+
``` python
36+
Segmentation.populate()
37+
```
38+
39+
=== "Matlab"
40+
41+
In Matlab, the `key` is a `struct`.
42+
``` matlab
43+
populate(Segmentation)
44+
```
45+
46+
For more information on the `populate` options in each language, please visit the
47+
[Python and Matlab](../../../) documentation pages.
48+
49+
<!-- TODO: Replace above with respective links when live -->

docs/src/compute/populate.md

Lines changed: 3 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,3 @@
1+
## Work in progress
2+
You may ask questions in the chat window below or
3+
refer to [legacy documentation](https://docs.datajoint.org/)

0 commit comments

Comments
 (0)