-
See https://github.com/lexibank/uralex
-
There are some mismappings, as they have like 6 words for DEER in the data. We were informed by somebody who wrote to Joshua Jackson, who then wrote to me:
```
Cow would be what is translated as D…
-
I read your paper with great interest and am happy you shared the code online. However, I realized that you probably do not know that we have dedicated Python APIs which you can conveniently use to ac…
-
For clics4, we will have some 52 datasets, all segmented and therefore analyzable with LingPy cognate detection methods. This means, we can offer enhanced networks (which require to integrate code tha…
-
We need to agree on a network representation format. The easiest way is to generate a graph with networkx and then to save it as GML, although this may be tricky at times. Another possibility is to ge…
-
@AnnaDiNatale and @katjabocklage, please help by expanding:
1. Does partial colexification (a in b-colexification) work as a good predictor of semantic change directions?
2. What other factors cou…
-
We are starting to work in the colexifications project. I think the best is to create a new repository colexifications-numeralbank and copying the numeralbank-analysed data in the new repository, is t…
-
Something along these lines:
```python
from collections import defaultdict
from xml.etree import ElementTree
import re
import io
import requests
from clldutils.source import Source
URL =…
-
In the sqlite database that is obtained after the CLICS install, what seems to be the same concept (parameter) is represented multiple times with different numeric identifiers. For example, consider t…
-
Add a reminder to the README that the analysis should ideally be run inside a virtual environment.