API: replace dropna=False option with na_sentinel=None in factorize #35852
Conversation
|
Hello @charlesdong1991! Thanks for updating this PR. We checked the lines you've touched for PEP 8 issues, and found: There are currently no PEP 8 issues detected in this Pull Request. Cheers! 🍻 Comment last updated at 2020-09-02 14:12:23 UTC |
charlesdong1991
changed the title
charlesdong1991
changed the title
charlesdong1991
added
Docs
Testing
charlesdong1991
changed the title
na_sentinel keyword in pd.factorize supports None to include NaNs in the unique of valuesna_sentinel keyword in pd.factorize supports None to include NaNs in the unique of values and remove dropna keyword
jorisvandenbossche
changed the title
na_sentinel keyword in pd.factorize supports None to include NaNs in the unique of values and remove dropna keyword
| If NaN is in the values, and we want to include NaN in the uniques of the | ||
| values, it can be achieved by setting ``na_sentinel=None``. | ||
|
|
||
| >>> values = np.array([1, 2, 1, np.nan]) |
There was a problem hiding this comment.
can you add the same example above this as well but w/o setting na_sentinel
pandas/core/algorithms.py
Outdated
| values, it can be achieved by setting ``na_sentinel=None``. | ||
|
|
||
| >>> values = np.array([1, 2, 1, np.nan]) | ||
| >>> codes, uniques = pd.factorize(values) # default |
There was a problem hiding this comment.
| >>> codes, uniques = pd.factorize(values) # default | |
| >>> codes, uniques = pd.factorize(values) # default na_sentinel=-1 |
| array([ 0, 1, 0, -1]) | ||
| >>> uniques | ||
| array([1., 2.]) | ||
| >>> codes, uniques = pd.factorize(values, na_sentinel=None) |
There was a problem hiding this comment.
ideally put a blank line as othewise this is very hard to read
|
@charlesdong1991 can ignore that 38_locale failure |
|
thanks @jreback Will ignore and merge it! |
|
@simonjayhawkins btw, i think should this be backported, right? |
|
@meeseeksdev backport 1.1.x |
…a_sentinel=None in factorize
yep. thanks. meeseeks... is only very occasionally picking these up now. |
|
i've got a branch going for automating the release process, but in the first instance am using it for release readiness, i'm planning to add a check shortly that the whatsnew on the backport and master are the same. so this should make sure that any PRs that change the whatsnew aren't missed. |
…=None in factorize (#36071) Co-authored-by: Kaiqi Dong <[email protected]>
That would be cool! |
…andas-dev#35852) * remove \n from docstring * fix issue 17038 * revert change * revert change * add dropna doc for factorize * rephrase the doc * flake8 * fixup * use NaN * add dropna in series.factorize * black * add test * linting * linting * doct * fix black * fixup * fix doctest * add whatsnew * linting * fix test * try one time * hide dropna and use na_sentinel=None * update whatsnew * rename test function * remove dropna from factorize * update doc * docstring * update doc * add comment * code change on review * update doc * code change on review * minor move in whatsnew * add default example * doc * one more try * explicit doc * add space
black pandasgit diff upstream/master -u -- "*.py" | flake8 --diff