Skip to content

Commit ca3a869

Browse files
m-mueckem-muecke
committed
Deploying to gh-pages from @ aa14a24 🚀
1 parent f490a37 commit ca3a869

File tree

7 files changed

+122
-116
lines changed

7 files changed

+122
-116
lines changed

‎index.html‎

Lines changed: 40 additions & 38 deletions
Some generated files are not rendered by default. Learn more about customizing how changed files appear on GitHub.

‎index.md‎

Lines changed: 39 additions & 37 deletions
Original file line numberDiff line numberDiff line change
@@ -72,20 +72,14 @@ prediction$score(msr("regr.rmse"))
7272

7373
# generate new data to forecast unseen data
7474
newdata = generate_newdata(task, 12L)
75-
newdata
76-
#> month passengers
77-
#> 1: 1961-01-01 NA
78-
#> 2: 1961-02-01 NA
79-
#> 3: 1961-03-01 NA
80-
#> 4: 1961-04-01 NA
81-
#> 5: 1961-05-01 NA
82-
#> 6: 1961-06-01 NA
83-
#> 7: 1961-07-01 NA
84-
#> 8: 1961-08-01 NA
85-
#> 9: 1961-09-01 NA
86-
#> 10: 1961-10-01 NA
87-
#> 11: 1961-11-01 NA
88-
#> 12: 1961-12-01 NA
75+
head(newdata)
76+
#> month passengers
77+
#> 1: 1961-01-01 NA
78+
#> 2: 1961-02-01 NA
79+
#> 3: 1961-03-01 NA
80+
#> 4: 1961-04-01 NA
81+
#> 5: 1961-05-01 NA
82+
#> 6: 1961-06-01 NA
8983
prediction = learner$predict_newdata(newdata, task)
9084
prediction
9185
#>
@@ -129,6 +123,14 @@ learner$predict_newdata(newdata, task)
129123
#> 10 NA 469.8626 474.5036 494.1275 513.7514 518.3925 494.1275 1961-10-01
130124
#> 11 NA 398.8383 403.5234 423.3336 443.1438 447.8290 423.3336 1961-11-01
131125
#> 12 NA 440.8230 445.5445 465.5085 485.4725 490.1940 465.5085 1961-12-01
126+
127+
# resampling
128+
learner = lrn("fcst.auto_arima")
129+
resampling = rsmp("fcst.holdout", ratio = 0.7)
130+
rr = resample(task, learner, resampling)
131+
rr$aggregate(msr("regr.rmse"))
132+
#> regr.rmse
133+
#> 27.1211
132134
```
133135

134136
### Example: forecasting with regression learner
@@ -145,39 +147,39 @@ prediction
145147
#>
146148
#> ── <PredictionRegr> for 12 observations: ───────────────────────────────────────
147149
#> row_ids truth response month
148-
#> 1 NA 436.3897 1961-01-01
149-
#> 2 NA 435.7855 1961-02-01
150-
#> 3 NA 457.3597 1961-03-01
150+
#> 1 NA 434.1089 1961-01-01
151+
#> 2 NA 435.1952 1961-02-01
152+
#> 3 NA 453.3213 1961-03-01
151153
#> --- --- --- ---
152-
#> 10 NA 474.5674 1961-10-01
153-
#> 11 NA 438.2343 1961-11-01
154-
#> 12 NA 440.4972 1961-12-01
154+
#> 10 NA 480.7029 1961-10-01
155+
#> 11 NA 440.3154 1961-11-01
156+
#> 12 NA 439.9652 1961-12-01
155157
prediction = flrn$predict(task, 140:144)
156158
prediction
157159
#>
158160
#> ── <PredictionRegr> for 5 observations: ────────────────────────────────────────
159161
#> row_ids truth response month
160-
#> 140 606 573.4641 1960-08-01
161-
#> 141 508 503.7301 1960-09-01
162-
#> 142 461 456.3900 1960-10-01
163-
#> 143 390 415.1462 1960-11-01
164-
#> 144 432 433.3205 1960-12-01
162+
#> 140 606 580.4945 1960-08-01
163+
#> 141 508 503.9283 1960-09-01
164+
#> 142 461 453.5133 1960-10-01
165+
#> 143 390 414.3402 1960-11-01
166+
#> 144 432 431.9185 1960-12-01
165167
prediction$score(msr("regr.rmse"))
166168
#> regr.rmse
167-
#> 18.61261
169+
#> 16.22105
168170

169171
flrn = as_learner_fcst(learner, lags = 1:12)
170172
resampling = rsmp("fcst.holdout", ratio = 0.9)
171173
rr = resample(task, flrn, resampling)
172174
rr$aggregate(msr("regr.rmse"))
173175
#> regr.rmse
174-
#> 46.53581
176+
#> 47.60131
175177

176178
resampling = rsmp("fcst.cv")
177179
rr = resample(task, flrn, resampling)
178180
rr$aggregate(msr("regr.rmse"))
179181
#> regr.rmse
180-
#> 26.69851
182+
#> 25.99407
181183
```
182184

183185
Or with some feature engineering using mlr3pipelines:
@@ -201,7 +203,7 @@ glrn = as_learner(graph %>>% flrn)$train(task)
201203
prediction = glrn$predict(task, 142:144)
202204
prediction$score(msr("regr.rmse"))
203205
#> regr.rmse
204-
#> 14.59454
206+
#> 14.55323
205207
```
206208

207209
### Example: forecasting electricity demand
@@ -228,13 +230,13 @@ prediction
228230
#>
229231
#> ── <PredictionRegr> for 14 observations: ───────────────────────────────────────
230232
#> row_ids truth response date
231-
#> 1 NA 186121.6 2015-01-01
232-
#> 2 NA 196014.4 2015-01-02
233-
#> 3 NA 189591.3 2015-01-03
233+
#> 1 NA 188367.1 2015-01-01
234+
#> 2 NA 197059.7 2015-01-02
235+
#> 3 NA 189557.6 2015-01-03
234236
#> --- --- --- ---
235-
#> 12 NA 221422.6 2015-01-12
236-
#> 13 NA 225356.3 2015-01-13
237-
#> 14 NA 226671.8 2015-01-14
237+
#> 12 NA 222577.0 2015-01-12
238+
#> 13 NA 227080.0 2015-01-13
239+
#> 14 NA 227389.6 2015-01-14
238240
```
239241

240242
### Example: global forecasting (longitudinal data)
@@ -267,14 +269,14 @@ flrn = as_learner_fcst(lrn("regr.ranger"), 1:3)$train(task)
267269
prediction = flrn$predict(task, 4460:4464)
268270
prediction$score(msr("regr.rmse"))
269271
#> regr.rmse
270-
#> 21157.43
272+
#> 20657.83
271273

272274
flrn = as_learner_fcst(lrn("regr.ranger"), 1:3)
273275
resampling = rsmp("fcst.holdout", ratio = 0.9)
274276
rr = resample(task, flrn, resampling)
275277
rr$aggregate(msr("regr.rmse"))
276278
#> regr.rmse
277-
#> 91069.37
279+
#> 91283
278280
```
279281

280282
### Example: global vs local forecasting

‎llms.txt‎

Lines changed: 39 additions & 37 deletions
Original file line numberDiff line numberDiff line change
@@ -72,20 +72,14 @@ prediction$score(msr("regr.rmse"))
7272

7373
# generate new data to forecast unseen data
7474
newdata = generate_newdata(task, 12L)
75-
newdata
76-
#> month passengers
77-
#> 1: 1961-01-01 NA
78-
#> 2: 1961-02-01 NA
79-
#> 3: 1961-03-01 NA
80-
#> 4: 1961-04-01 NA
81-
#> 5: 1961-05-01 NA
82-
#> 6: 1961-06-01 NA
83-
#> 7: 1961-07-01 NA
84-
#> 8: 1961-08-01 NA
85-
#> 9: 1961-09-01 NA
86-
#> 10: 1961-10-01 NA
87-
#> 11: 1961-11-01 NA
88-
#> 12: 1961-12-01 NA
75+
head(newdata)
76+
#> month passengers
77+
#> 1: 1961-01-01 NA
78+
#> 2: 1961-02-01 NA
79+
#> 3: 1961-03-01 NA
80+
#> 4: 1961-04-01 NA
81+
#> 5: 1961-05-01 NA
82+
#> 6: 1961-06-01 NA
8983
prediction = learner$predict_newdata(newdata, task)
9084
prediction
9185
#>
@@ -129,6 +123,14 @@ learner$predict_newdata(newdata, task)
129123
#> 10 NA 469.8626 474.5036 494.1275 513.7514 518.3925 494.1275 1961-10-01
130124
#> 11 NA 398.8383 403.5234 423.3336 443.1438 447.8290 423.3336 1961-11-01
131125
#> 12 NA 440.8230 445.5445 465.5085 485.4725 490.1940 465.5085 1961-12-01
126+
127+
# resampling
128+
learner = lrn("fcst.auto_arima")
129+
resampling = rsmp("fcst.holdout", ratio = 0.7)
130+
rr = resample(task, learner, resampling)
131+
rr$aggregate(msr("regr.rmse"))
132+
#> regr.rmse
133+
#> 27.1211
132134
```
133135

134136
### Example: forecasting with regression learner
@@ -145,39 +147,39 @@ prediction
145147
#>
146148
#> ── <PredictionRegr> for 12 observations: ───────────────────────────────────────
147149
#> row_ids truth response month
148-
#> 1 NA 436.3897 1961-01-01
149-
#> 2 NA 435.7855 1961-02-01
150-
#> 3 NA 457.3597 1961-03-01
150+
#> 1 NA 434.1089 1961-01-01
151+
#> 2 NA 435.1952 1961-02-01
152+
#> 3 NA 453.3213 1961-03-01
151153
#> --- --- --- ---
152-
#> 10 NA 474.5674 1961-10-01
153-
#> 11 NA 438.2343 1961-11-01
154-
#> 12 NA 440.4972 1961-12-01
154+
#> 10 NA 480.7029 1961-10-01
155+
#> 11 NA 440.3154 1961-11-01
156+
#> 12 NA 439.9652 1961-12-01
155157
prediction = flrn$predict(task, 140:144)
156158
prediction
157159
#>
158160
#> ── <PredictionRegr> for 5 observations: ────────────────────────────────────────
159161
#> row_ids truth response month
160-
#> 140 606 573.4641 1960-08-01
161-
#> 141 508 503.7301 1960-09-01
162-
#> 142 461 456.3900 1960-10-01
163-
#> 143 390 415.1462 1960-11-01
164-
#> 144 432 433.3205 1960-12-01
162+
#> 140 606 580.4945 1960-08-01
163+
#> 141 508 503.9283 1960-09-01
164+
#> 142 461 453.5133 1960-10-01
165+
#> 143 390 414.3402 1960-11-01
166+
#> 144 432 431.9185 1960-12-01
165167
prediction$score(msr("regr.rmse"))
166168
#> regr.rmse
167-
#> 18.61261
169+
#> 16.22105
168170

169171
flrn = as_learner_fcst(learner, lags = 1:12)
170172
resampling = rsmp("fcst.holdout", ratio = 0.9)
171173
rr = resample(task, flrn, resampling)
172174
rr$aggregate(msr("regr.rmse"))
173175
#> regr.rmse
174-
#> 46.53581
176+
#> 47.60131
175177

176178
resampling = rsmp("fcst.cv")
177179
rr = resample(task, flrn, resampling)
178180
rr$aggregate(msr("regr.rmse"))
179181
#> regr.rmse
180-
#> 26.69851
182+
#> 25.99407
181183
```
182184

183185
Or with some feature engineering using mlr3pipelines:
@@ -201,7 +203,7 @@ glrn = as_learner(graph %>>% flrn)$train(task)
201203
prediction = glrn$predict(task, 142:144)
202204
prediction$score(msr("regr.rmse"))
203205
#> regr.rmse
204-
#> 14.59454
206+
#> 14.55323
205207
```
206208

207209
### Example: forecasting electricity demand
@@ -228,13 +230,13 @@ prediction
228230
#>
229231
#> ── <PredictionRegr> for 14 observations: ───────────────────────────────────────
230232
#> row_ids truth response date
231-
#> 1 NA 186121.6 2015-01-01
232-
#> 2 NA 196014.4 2015-01-02
233-
#> 3 NA 189591.3 2015-01-03
233+
#> 1 NA 188367.1 2015-01-01
234+
#> 2 NA 197059.7 2015-01-02
235+
#> 3 NA 189557.6 2015-01-03
234236
#> --- --- --- ---
235-
#> 12 NA 221422.6 2015-01-12
236-
#> 13 NA 225356.3 2015-01-13
237-
#> 14 NA 226671.8 2015-01-14
237+
#> 12 NA 222577.0 2015-01-12
238+
#> 13 NA 227080.0 2015-01-13
239+
#> 14 NA 227389.6 2015-01-14
238240
```
239241

240242
### Example: global forecasting (longitudinal data)
@@ -267,14 +269,14 @@ flrn = as_learner_fcst(lrn("regr.ranger"), 1:3)$train(task)
267269
prediction = flrn$predict(task, 4460:4464)
268270
prediction$score(msr("regr.rmse"))
269271
#> regr.rmse
270-
#> 21157.43
272+
#> 20657.83
271273

272274
flrn = as_learner_fcst(lrn("regr.ranger"), 1:3)
273275
resampling = rsmp("fcst.holdout", ratio = 0.9)
274276
rr = resample(task, flrn, resampling)
275277
rr$aggregate(msr("regr.rmse"))
276278
#> regr.rmse
277-
#> 91069.37
279+
#> 91283
278280
```
279281

280282
### Example: global vs local forecasting

‎pkgdown.yml‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -2,7 +2,7 @@ pandoc: 3.1.11
22
pkgdown: 2.2.0
33
pkgdown_sha: ~
44
articles: {}
5-
last_built: 2025-12-03T15:32Z
5+
last_built: 2025-12-03T20:51Z
66
urls:
77
reference: https://mlr3forecast.mlr-org.com/reference
88
article: https://mlr3forecast.mlr-org.com/articles

‎reference/mlr_learners_fcst.auto_adam.html‎

Lines changed: 1 addition & 1 deletion
Some generated files are not rendered by default. Learn more about customizing how changed files appear on GitHub.

‎reference/mlr_learners_fcst.auto_adam.md‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -202,7 +202,7 @@ learner$train(task, row_ids = ids$train)
202202

203203
# Print the model
204204
print(learner$model)
205-
#> Time elapsed: 4.04 seconds
205+
#> Time elapsed: 4.01 seconds
206206
#> Model estimated using auto.adam() function: ETS(MAM)
207207
#> With backcasting initialisation
208208
#> Distribution assumed in the model: Normal

‎search.json‎

Lines changed: 1 addition & 1 deletion
Large diffs are not rendered by default.

0 commit comments

Comments
 (0)