forked from pveber/morse
-
Notifications
You must be signed in to change notification settings - Fork 0
/
NEWS
256 lines (201 loc) · 9.87 KB
/
NEWS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
=== morse : MOdelling tools for Reproduction and Survival data in Ecotoxicology ===
Version 3.3.1
=============
FIX ISSUE Github #286 Accomodate new value for default.stringAsFactors()
- replicate in survData are now only of class factor
Version 3.3.0
=============
IMPORTANT NOTE
- This release should mainly work with all objects created with earlier version >3.0.0.
- However, in version 3.2.7 we introduced 'hb_valueFIXED' (in survFit) and 'hb_valueFORCED' (in prediction) to play with background mortality (hb_value). You can still use the functions (as MFx) but make sure to provide appropriate parameters to 'hb_valueFIXED' and 'hb_valueFORCED' (see package documentation for details).
NEW SOLVER
- a new ODE solver has been introduced reducing the time of computation in prediction function ('predict', 'predict_Nsurv' and 'MFx').
NEW NSURV SAMPLER
- a new sampler for prediction of Nsurv is introduced. The speed of the function 'predict_Nsurv' should be 30 to 60 times faster.
FIX ISSUE
- add 'interpolate_length' option to MFx to release the constraint at 100 time points.
ADD DATA
- add FOCUS profile to the data set for testing.
DOCUMENTATION
- improve links between documentation of functions (i.e., join documentation of similar functions).
- add vignette for a minimal working example.
Version 3.2.7
=============
TYPO FIX
- add 'hb_valueFIXED' (in survFit) and 'hb_valueFORCED' (in prediction) to fix hb_value when hb_value is FALSE
FIX WARNING MESSAGES
- fix bug returning warning messages about depecration of "as_data_frame" and "data_frame", replaced respectivelly by "as_tibble" and "tiblle"
Version 3.2.6
=============
TYPO FIX
- correct 'alpha' in 'mw' in summary of survFit
Version 3.2.5
=============
TYPO FIX
- change 'survival rate' into 'survival probability' in plots
Version 3.2.4
=============
DOC ISSUES
- update co-authors
- fix some typos in documents
Version 3.2.3 (no submitted to CRAN)
=============
BUG FIXES
- SQRT was not applied to NRMSE, so it was a NMSE.
Version 3.2.2
=============
OTHER CHANGES
- add warning message in 'survFit' function when 'hb_value' is fixed.
- add function 'predict_Nsurv_ode' function to compute predictive function with
the ODE solver from deSolve package.
Version 3.2.1
=============
DOC ISSUES FIXES
- changes in the vignette 'Tutorial'.
- changes in the Reference manual 'morse.pdf'.
NEW FEATURES
- Details NRMSE and PPC for each time-series (i.e., replicates)
Version 3.2.0
=============
BUG FIXES
- Change default value for argument 'adddata' in function 'plot' as TRUE
NEW FEATURES
- add 'predict_Nsurv' function to compute prediction for the number of survivors.
- add 'predict_Nsurv_check' function to compute criteria to check model.
performance: (i) percentage of observation within the 95% credible interval of
prediction drawn from posterior, (ii) the Normalised Root Mean Square Error
(NRMSE) and (iii) the Survival-Probability Prediction Error (SPPE).
- add 'ppc.survFitPredict_Nsurv' function to plot ppc for 'survFitPredict_Nsurv'
object
- add 'plot.survFitPredict_Nsurv' function to plot 'survFitPredict_Nsurv' object
- add 'Nsurv_ppc' and 'Nsurv_sim' in 'mcmc' object within 'survFit' class objects
- Modif JAGS models to ensure probability use for binomial distribution is <1 (JAGS does not work with p=1).
- add option to remove background mortality, 'hb_value' TRUE/FALSE, in 'survFit' methods.
- add 'predict_ode' function to include deSolve package for ode integration for prediction.
- add option 'EFSA_name' in function 'summary', 'plot_prior_posterior' and 'priors_distribution' to adapt parameter name of GUTS model to EFSA Scientific Opinion (doi: 10.2903/j.efsa.2018.5377) naming.
- add binomial credible interval of observation in plot for 'survFitVarExp' object
Version 3.1.1
=============
DOC ISSUES FIXES
- changes in the tutorial vignettes.
Version 3.1.0
=============
NEW FEATURES
- add 'priors_distribution' function to easly returned the distribution of priors.
- add 'MFx' function to compute the Multiplication Factor leading to a
reduction of x percent of survival. Return an object of class 'MFx'.
- add 'plot.MFx' function to plot object of class 'MFx'.
BUG FIXES
- change the automatized title of the plot of an 'LCx' object.
Version 3.0.0
=============
NEW FEATURES
- add survival dataset for pulse exposure: 'propiconazole_pulse_exposure'.
- add S3 'survFit' method for TKTD modeling with model SD or IT for both 'time constant exposure' or 'time varying exposure'.
- add 'LCx' function to plot Lethal Concentration for x percent of the population for survFit model (TKTD SD or IT with constant or varying exposure).
- add option 'main' to add title in plot for PPC.
- add section for 'survFit' TKTD IT and SD (cst and var exposure) in the tutorial vignette.
BUG FIXES
- change the require dataset for survData and reproData: replicates are unique for each time-serie all over concentration profiles
- change the structure of all datasets included to satisfy that all replicates are unique for each time-serie all over concentration profiles.
Version 2.2.0
=============
NEW FEATURES
- add the 'survFitTKTD' function to analyse 'survData' object with a time-exposur response model.
- add the S3 'plot', 'print', 'summary' and 'ppc' methods for the 'survFitTKTD'
class.
- add two survival datasets : dichromate and propiconazole.
- the method plotDoseResponse replaces the plot.survData and plot.reproData
when the argument target.time was set. Now the response is expressed in
survival rate or in number of offspring per individual-days.
- the calcul of the mean fitted curve of 'plot.reproFitTT' and 'plot.survFitTT'
is the same of the 95 % Credible limits.
- the ci arg in 'plot.reproFitTT' and 'plot.survFitTT' was suppressed and the
credible limites was allways plotted.
- added a new argument 'adddata' in 'plot.reproFitTT' and 'plot.survFitTT'
to draw the datas with theyre confidence interval.
- add a new argument 'spaghetti' in 'plot.reproFitTT' and 'plot.survFitTT'
to draw the credible interval as multiple curves.
Version 2.1.1
=============
BUG FIXES
- update the cross-references link to ggplot2 package
Version 2.1.0
=============
SYSTEM REQUIREMENTS
- JAGS version 4.0.0 or higher is required
- rjags version 4.4 or higher is required
NEW FEATURES
- the dataset cadmium2 now reports the number of eggs instead of the number of
clutches.
- add S3 method ppc to plot the posterior predictive check on reproFitTT and
survFitTT objects.
- added a new argument 'remove.someLabels' to plot.reproData and plot.survData
functions to avoid the overlap of labels of the X-axis in ggplot style.
BUG FIXES
- the check on replicate's label has been removed in the survDatacheck
function.
- minor changes in the tutorial vignette.
- remove titles in legend of plot.reproFitTT and plot.survFitTT in
ggplot style.
- remove observed value points in plot.reproFitTT.
- survDataCheck and reproDataCheck would fail on datasets with integer-only
concentrations.
- the summary functions for survData and reproData now don't show the
number of datapoints per concentration and the number of datapoints
per time. This was considered misleading in the case where some
replicates have a different initial number of individuals.
- plot.survData (and plot.reproData) don't display a legend anymore with
unconstrained time and concentration. They used to display a misleading
"replicate" legend, which would not correspond to a real replication unit
(beaker).
- in plot.survData all ggplot style graphique start at 0,0.
OTHER CHANGES
- the DIC is not calculated in survFiTT anymore. Since we don't
perform model selection, this was not really useful and we thus
obtain a significant speed-up.
- the column ID was removed in survData and reproData objects.
Version 2.0.0
=============
NEW FEATURES
- new package structure and function names.
- the repro.data function is now called reproData and creates an object of class
reproData without specific target time.
- the repro.check.data function is now called reproDataCheck.
- the repro.fullsurvplot and repro.cumulplot functions are replaced by the generic
plot S3 method for the class reproData.
- the repro.survplot function now depends on the plot S3 metod for the new class
survData.
- the repro.fit function is now called reproFitTT with the associated class
reproFitTT, with the argument target.time.
- add of new arguments for the reproFitTT function:
- "stoc.part" to choose the stochastic part of the model.
- "ecx" to choose the value of ECx estimated by the function.
- the deterministic part name are now in the legend of the plot function.
- when stoc.part = "bestfit" is choosen, the model "gammapoisson" is chosen if
the DIC of gamma-poisson model is smaller than DIC of poisson model of 10
(1 in the previous version).
- upgrade the informations in the summary of the reproData class.
- upgrade the informations in the summary of the reproFitTT function, and values
of parameters are expressed in untransformed values.
- the credible interval in plot.reproFitTT are now a ribbon.
- add a new group of functions to analyze survival data:
- the survDataCheck function checks the integrity of the dataset.
- the survData function with associated S3 method creates survData object.
- the survFitTT function run the bayesian estimation to fit an exposure-response
model for survival data with the associated specific S3 method: print, plot
and summary.
- the lattice graphics was suppressed.
- the repro.convergence function was suppressed.
BUGS FIXES
- new selection rules for the "bestfit" method in the function reproFitTT.
- in log scale the points at concentration = 0 are not drawn (ggplot).
Version 1.0.2
=============
BUG FIXES
- correction of points color in the legend for the plot.repro.fit function (ggplot).
- change the ylim range for the plot.repro.fit function (generic).
- correction of the log.scale option for the repro.survplot function.
Version 1.0.1
=============
- Initial release.