|
1 | | -# inphr (development version) |
| 1 | +# inphr 0.0.1 |
2 | 2 |
|
3 | | -* Initial CRAN submission. |
| 3 | +This is a new submission to CRAN. |
| 4 | + |
| 5 | +## Goals |
| 6 | + |
| 7 | +The [{inphr}](https://tdaverse.github.io/inphr/) package is intended to be a |
| 8 | +package for making inference on samples of persistence homology data. It is |
| 9 | +part of the *TDAverse* suite of packages, which are designed to provide a |
| 10 | +collection of packages for enabling machine learning and data science tasks |
| 11 | +using persistent homology. |
| 12 | + |
| 13 | +## Current features |
| 14 | + |
| 15 | +The package currently exposes two main functions which test if two samples of PH data have been generated from the same distribution: |
| 16 | + |
| 17 | +- [`two_sample_diagram_test()`](https://tdaverse.github.io/inphr/reference/two_sample_diagram_test.html) works in the space of diagrams, using test statistics based on inter-point distances only. |
| 18 | +- [`two_sample_functional_test()`](https://tdaverse.github.io/inphr/reference/two_sample_functional_test.html) works in a functional space (one of Betti, Euler characteristic, normalized life, silhouette or entropy) and uses interval-wise testing (providing strong control of familywise error rate) to output on which portions of the scale sequence does the difference occur. |
| 19 | + |
| 20 | +## Dependencies |
| 21 | + |
| 22 | +Messages, warnings and errors are relayed to the user using the |
| 23 | +[{rlang}](https://rlang.r-lib.org) package and the |
| 24 | +[{cli}](https://cli.r-lib.org) package which are both licensed under the MIT |
| 25 | +license and with no dependency trail. |
| 26 | + |
| 27 | +Inference in the space of diagrams is performed thanks to the combination of |
| 28 | +[{phutil}](https://cran.r-project.org/package=phutil) which computes distances between diagrams in an efficient manner and |
| 29 | +[{flipr}](https://cran.r-project.org/package=flipr) which powers the permutation schemes and test statistics based on |
| 30 | +inter-point distances. |
| 31 | + |
| 32 | +Inference in functional spaces is performed thanks to the combination of |
| 33 | +[{TDAvec}](https://cran.r-project.org/package=TDAvec) which provides the suitable PH vectorization and [{fdatest}](https://cran.r-project.org/package=fdatest) which |
| 34 | +powers the interval-wise testing procedure for functional data. |
0 commit comments