You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Suppose input values are defined on [5, 14], Then if you use lowess and use xvals to get values on a bigger range of Xs (for example [0, 19]), the output may contain undefined values even on the input range of Xs. Additionally, different frac values produce different behaviours that are not obvious.
In the last example, output is undefined even if you use the same output X values as input X values.
Ideally, lowess output should be defined at least on the range of input Xs.
Also, values do not become NaNs gradually (for example undefined below some frac value and defined above).
Because of that it looks like a bug and not like an expected behaviour.
Output of import statsmodels.api as sm; sm.show_versions()
INSTALLED VERSIONS
------------------
Python: 3.6.10.final.0
OS: Linux 5.4.95-1-MANJARO #1 SMP PREEMPT Thu Feb 4 00:38:02 UTC 2021 x86_64
byteorder: little
LC_ALL: None
LANG: en_US.UTF-8
statsmodels
===========
Installed: 0.12.2 (/home/timofey/.miniconda3/lib/python3.6/site-packages/statsmodels)
Required Dependencies
=====================
cython: 0.29.21 (/home/timofey/.miniconda3/lib/python3.6/site-packages/Cython)
numpy: 1.19.1 (/home/timofey/.miniconda3/lib/python3.6/site-packages/numpy)
scipy: 1.5.2 (/home/timofey/.miniconda3/lib/python3.6/site-packages/scipy)
pandas: 1.1.5 (/home/timofey/.miniconda3/lib/python3.6/site-packages/pandas)
dateutil: 2.8.1 (/home/timofey/.miniconda3/lib/python3.6/site-packages/dateutil)
patsy: 0.5.1 (/home/timofey/.miniconda3/lib/python3.6/site-packages/patsy)
Optional Dependencies
=====================
matplotlib: Not installed
cvxopt: Not installed
joblib: Not installed
Developer Tools
================
IPython: 7.16.1 (/home/timofey/.miniconda3/lib/python3.6/site-packages/IPython)
jinja2: 2.11.3 (/home/timofey/.miniconda3/lib/python3.6/site-packages/jinja2)
sphinx: Not installed
pygments: 2.7.4 (/home/timofey/.miniconda3/lib/python3.6/site-packages/pygments)
pytest: Not installed
virtualenv: 16.7.5 (/home/timofey/.miniconda3/lib/python3.6/site-packages)
The text was updated successfully, but these errors were encountered:
tprodanov
changed the title
Lowess extrapolation missing output values.
BUG: Lowess extrapolation missing output values.
Feb 26, 2021
Bug description
Suppose input values are defined on [5, 14], Then if you use
lowess
and usexvals
to get values on a bigger range of Xs (for example [0, 19]), the output may contain undefined values even on the input range of Xs. Additionally, differentfrac
values produce different behaviours that are not obvious.In the last example, output is undefined even if you use the same output X values as input X values.
Code example
Expected output
Ideally,
lowess
output should be defined at least on the range of input Xs.Also, values do not become NaNs gradually (for example undefined below some
frac
value and defined above).Because of that it looks like a bug and not like an expected behaviour.
Output of
import statsmodels.api as sm; sm.show_versions()
The text was updated successfully, but these errors were encountered: