Closed LucLapenta closed 4 years ago
Merging #276 into master will increase coverage by
0.09%
. The diff coverage isn/a
.
@@ Coverage Diff @@
## master #276 +/- ##
==========================================
+ Coverage 95.12% 95.21% +0.09%
==========================================
Files 22 22
Lines 3178 3178
==========================================
+ Hits 3023 3026 +3
+ Misses 155 152 -3
Impacted Files | Coverage Δ | |
---|---|---|
pygam/pygam.py | 94.79% <0.00%> (+0.12%) |
:arrow_up: |
pygam/utils.py | 87.73% <0.00%> (+0.30%) |
:arrow_up: |
pygam/tests/test_GAM_methods.py | 100.00% <0.00%> (+0.36%) |
:arrow_up: |
Continue to review full report at Codecov.
Legend - Click here to learn more
Δ = absolute <relative> (impact)
,ø = not affected
,? = missing data
Powered by Codecov. Last update 2323eb7...5e2f21f. Read the comment docs.
@LucLapenta Thanks for the PR! looks good :)
Updated documentation reference code on tuning a linearGAM model to be accurate compared to what the comments describe.
In the original reference code the numerical outputs at each step did not match the expected results based on the comments:
lams = np.random.rand(100, 3) # random points on [0, 1], with shape (100, 3)
lams = lams * 8 - 3 # shift values to -3, 3
lams = np.exp(lams) # transforms values to 1e-3, 1e3
The second step
lams = lams * 8 - 3 # shift values to -3, 3
results in shifting the random values from -3,5 rather than -3,3.The third step
lams = np.exp(lams) # transforms values to 1e-3, 1e3
is computing e^lams when is seems like it is supposed to be computing 10^lams.After running the original reference code, the min and max values are the resulting matrix are 0.04992659817158999 and 148.3540116029961 respectively.
The reference code in the quick_start.ipynb notebook was modified to:
lams = np.random.rand(100, 3) # random points on [0, 1], with shape (100, 3)
lams = lams * 6 - 3 # shift values to -3, 3
lams = 10 ** lams # transforms values to 1e-3, 1e3
Which results in more correct minimum and maximum values in the range of [1e-3, 1e3].