-
Notifications
You must be signed in to change notification settings - Fork 1
/
CHANGES.txt
270 lines (231 loc) · 9.6 KB
/
CHANGES.txt
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
257
258
259
260
261
262
263
264
265
266
267
268
269
270
v0.4.10, 10 November 2016
Fixed Bugs:
1. Fixed issue in PlotDist that prevented plotting
RVs with numbers in the support that aren't standard
python types
v0.4.9, 09 November 2016
Fixed Bugs:
1. Fixed indexing problem in transform procedure
was previously ignoring some segments of the transform
v0.4.8, 06 November 2016
New Capabilities:
Fixed Bugs:
1. Minor change to setup.py
v0.4.7, 06 November 2016
New Capabilities:
Fixed Bugs:
1. Corrected Convolution function (accidentally broke it
when [0,1] shortcut was implemented for convolution)
2. ArcTanRV was corrected
v0.4.5, 22 October 2016
New Capabilities:
1. Implemented shortcut for convolutions of two
random variables with [0,1] support
Fixed Bug:
1. Removed pylab from the setup.py file. This is installed
with matplotlib
2. Fixed bug in posterior. Does integration directly instead
of through Truncate when the support of an RV has oo or -oo
v0.4.4, 20 August 2016
New Capabilities:
1. PlotLimits implemented to change the limits of a plot
Fixed Bugs:
1. Fixed an issue in IDF where output wasn't returned
for continuous distributions
v0.4.3, 19 July 2016
New Capabilities:
1. X.save() method implemented to stored random variables
as binary
2. LoadRV() function implemented to load random variables
from binary
v0.4.2, 12 June 2016
New Capabilities:
1. New style plots
Fixed Bugs:
1. Added code that allows users to specify variate
in X.variate for NormalRV
Other Comments:
1. Removed the plot.py file. Plotting no longer depends on it
2. setup.py file has been updated. pip install now downloads
all dependencies
v0.4.1,
New Capabilities:
Fixed Bugs:
1. Imported atan function in dist_type to allow for proper
initialization of the ArcTanRV
2. Fixed Transform to work with float fractions in the
transformation
3. Fixed bug in Product that didn't allow for the use of
rational numbers as parameters
4. Fixed CDF and SF so that values passed outside of the
support are returned as 0 or 1 instead of returning
an error
5. If Convolution or Product cannot compute, APPLPy is now
programmed to try to reverse the order of the
calculation before returning and error
6. Assumptions added to the following some distributions to
make computation of CDF more feasible
7. Truncate modified to divide by the correct normalizing
constant
8. SF discrete was modified the handle intermediate values
9. Added 'shortcut' for Weibull CDF
v0.4.0, 17 APR 2016
New Capabilities:
1. Discrete Time Markov Chains module
2. Improved method implmented for converting discrete RVs between
different functional forms. Now runs more efficiently
3. Discrete RVs with explicit values and discrete RVs defined by
expressions can now be manipulated with RV algebra
4. Sqrt(X) syntax implemented for transformation by sqrt(x).
5. X**n syntax implemented for transformations by x**n
6. Markov Chain class added, includes procedures for computing
the state of a system after n additional steps
7. Support added for rational numbers in DiscreteUniformRV
8. Support for discrete random variables added for the Expected
Value function
9. BernoulliRV type added
10. RangeStat procedure added for finding the range stastics for
a random variable
11. Maximum and Minimum IID now work with symbolic n unspecified
12. Random Variable method added to simplify functions with
assumptions
13. VerifyPDF(X) added for compatiblity with APPL
Fixed Bugs:
1. BootstrapRV procedure fixed. Duplicate elements are now
eliminated from the support of Bootstrap RVs.
2. Pr(X) for elements in the BootstrapRV are now described by
rational numbers instead of floats.
3. Bug in transform for RV's with negative support is fixed.
Transform is now running properly for all supports.
4. Fixed bug in convolution for lifetime distributions. When
a floating point value was entered for a parameter,
the procedure would return 0 for the cdf of the convolution.
The procedure now works correctly.
5. Added assumptions to variables in product procedure. No longer
gives complex functions as output for products that occur in
the third quadrant.
6. Corrections made to transform for discrete random variables. No
longer cuts out parts of the distribution that are not in
the support of the tranformation, but simply does not
transform them.
7. Corrections made to Transform to account for equality testing with
SymPy number types
8. Corrections made to the discrete transform procedure to handle
situations in which the support of the transformation does
not cover the full support of the random variable
9. Added simplification line to end of Truncate procedure
10. Fixed bug in computing values for discrete pdf
11. Fixed bug related to floating point numbers in transform
12. Fixed bug to correct computation of IDFs for Discrete RVs
13. OrderStat modified to allow for 'wo' for samples of size
greater than 4
14. Transform modified to accept symbolic discrete RVs
15. Convolution now works with rational supports
16. IDF for symbolic discrete random variables updated to
start the support at 0
17. Corrected likelihood function computation in Posterior
18. Fixed technique for finding area in Truncate
19. ChiRV now outputs rational numbers
20. Transform modified to handle transformations by 1/x and log(x)
v0.3.0, 12 JAN 2015
New Capabilities:
1. The Queue procedure computes the sojourn time distribution
for an M/M/s Queue for customer n, given k customers
initially in the queue.
2. Special MLE algorithms implemented for the Normal, Exponential,
Poisson and Weibull distributions
3. Thorough doc strings added for each applpy module
4. Cache option has been added for functional form conversion and
expected value procedure. If cache=True, APPLPy stores the
values in memory to be retrieve later if the function is
called again. This save significant CPU time on future
function calls
5. Doc Strings added for each distribution type
6. Behavior of +,-,*,/ set to allow random variables to interact
with floats and ints
7. == operator will now test for the equality of two random variables
8. Maximum and Minimum will now accept a variable number of arguments
9. Box-Muller Transformation method implemented to allow for more
efficient computation of Normal variates
10. Users can now specify method for variate generation (generally
inverse cdf or a numerical method)
11. X.latex() command added to generate the LaTeX code for a
random variable
Fixed Bugs:
1. Maximum and Minimum are now working properly for piecewise
distributions. Boundary conditions caused to the procedures
to use the incorrect segments in the earlier
implementation
2. The gamma function is now loaded into the namespace in the
dist_rv module.
3. Error message added to Transform to indicate if the user has
specified an undefined transformation
v0.2.1, 18 NOV 2014
New Capabilities:
1. Pre-processing code has been added to the beginning
of several procedures. Errors will be raised
if a procedure receives bad input.
2. PlotDist has been improved to allow for more plotting
options. PlotDist is now capable of using different
colors and displaying multiple plots simultaneously
3. Discrete and continuous plots can now be displayed in
the same overlay
4. Improved display implemented for continuous random variables.
Displays large piecewise random variabls more cleanly
5. MLE and MOM now have options for numerical solving
6. Procedure added for computing the entropy of random variables
Fixed Bugs:
1. Start-up script has been modified. The init_session()
commands for SymPy are entered manually. APPLPy
can now be imported in an interactive command
line session
2. Updates have been made to the way the APPLPy package is
structured. Importing APPLPy is now less memory
instensive (does not import all of SymPy, etc)
3. KSRV now runs properly ... arrays are no initialized
with 0 instead of None to avoid TypeErrors
when variables are summed
4. Minimum and Maximum procedures have been debugged for
discrete random variables. Old version had issues
recognizing floating point numbers in the support
v0.2.0, 17 SEPTEMBER 2014
New Capabilities:
1. Bayesian statistics module has been added. Capabilities
include:
a. Computing posterior distributions
b. Computing posterior predictive distributions
c. Computing Jeffreys Priors
d. Credible Sets
2. Functionality added for discrete random variables in
functional forms for the following procedures:
a. All procedures for changing functional form
(i.e. CDF to PDF)
b. All expected value procedures
3. New Random Variable types added (new distributions are
listed in the startup menu)
4. KS Test added to the statistics module
Fixed Bugs:
1. PlotDist now supports univariate discrete random
variables
2. Random Variables no longer default to floating point
numbers when it produces output. Computation of
exact distributions is now possible
3. PlotDist now appears in the iPython Notebook when the
command %pylab inline is entered
v0.1.2, 30 JULY 2014
New Capabilities:
1. Capability to work with RV's with symbolic supports was
added.
v0.1.1, 29 JULY 2014
Fixed Bugs:
1. Convolution/Product/MaximumIID procedures have been updated
so that they now function with all distributions. In the
initial release, they only worked with lifetime distributions.
2. Expected value procedures were modified to output
simplified expressions
3. init.py file modified so that all dependencies are
automatically imported
4. Plotting now takes place through the sympy plotting front-
end. The PlotDist procedure is now capable of plotting all
functions
v0.1.0, 23 JULY 2014 -- Initial Release