Add examples of different ways of creating string datasets #2424
Conversation
| string_data = ["varying", "sizes", "of", "strings"] | ||
|
|
||
| # Variable length strings (implicit) | ||
| f['vlen_strings1'] = string_data |
There was a problem hiding this comment.
This f is supposed to be the database? What is supposed to happen when you are doing it implicitly?
There was a problem hiding this comment.
f for file - this is a convention across the h5py docs, e.g. look further down this page, or at the dataset page. We don't really use the word database at all; HDF5 also talks about files rather than databases.
'Implicit' here just means you're not telling h5py a dtype, so it's guessing based on the object you give it.
| f['vlen_strings1'] = string_data | ||
|
|
||
| # Variable length strings (explicit) | ||
| ds = f.create_dataset('vlen_strings2', shape=4, dtype=h5py.string_dtype()) |
There was a problem hiding this comment.
It'd be nice if there were some examples with data= keyword argument - those are the ones that got me confused.
There was a problem hiding this comment.
Passing data= should be equivalent to the implicit case (f[name] =) if you don't pass dtype=, or to the explicit case if you do. I'm showing the explicit dtype case as two lines to illustrate that it lets you create the dataset before you have all the data.
I'd rather not make this example longer by showing more possible ways to do the same thing. The explanation at creating datasets could probably be improved as well.
| You can use :func:`.string_dtype` to explicitly specify any HDF5 string datatype. | ||
|
|
||
| :: |
There was a problem hiding this comment.
| You can use :func:`.string_dtype` to explicitly specify any HDF5 string datatype. | |
| :: | |
| You can use :func:`.string_dtype` to explicitly specify any HDF5 string datatype, | |
| as shown in the examples below:: |
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files@@ Coverage Diff @@
## master #2424 +/- ##
=======================================
Coverage 89.58% 89.58%
=======================================
Files 17 17
Lines 2391 2391
=======================================
Hits 2142 2142
Misses 249 249 ☔ View full report in Codecov by Sentry. |
Building on #2423, trying to illustrate the different possible ways to create string datasets.