provided up-to-date todo list to answer #1

readme.md

@@ -21,6 +21,8 @@ This implementation lets you:
 - predict the mean and variance and covariance matrix for given inputs
 - sample the distribution at a given position
 
+(See the [todo](https://github.com/nestordemeure/friedrich/blob/master/todo.md) file to get up-to-date informations on current developements.)
+
 ## Code sample
 
 ```rust
@@ -59,17 +61,6 @@ Most methods of this library can currently work with the following `input -> oup
 
 A trait is provided to add your own pairs.
 
-## Future developements
-
-The list of things that could be done to improve on the current implementation includes:
-
-- Reduce memory consumption for large datasets.
-- Add [ndarray](https://docs.rs/ndarray/) support behind a feature flag.
-- Add simple [kernel regression](https://en.wikipedia.org/wiki/Kernel_regression#Nadaraya%E2%80%93Watson_kernel_regression) (not as clever but much faster).
-- Add scalable variations of the algorithm.
-
-*Do not hesitate to send pull request or open issues to ask for features.*
-
 ## Why call it Friedrich ?
 
 Gaussian Process are named after the [Gaussian distribution](https://en.wikipedia.org/wiki/Gaussian_function) which is itself named after [Carl Friedrich Gauss](https://en.wikipedia.org/wiki/Carl_Friedrich_Gauss).

todo.md

@@ -1,6 +1,18 @@
-# TODO list
+# TODO
 
-- how to know if a kernel is default defined or user defined ?
-- add [benchmark](http://www.resibots.eu/limbo/release-2.0/reg_benchmarks.html)
-- integrate all nalgebra fix once new version is released
-- introduce a macro (to have optional parameters with default value) instead of the builder pattern ? can it let us deal with knowing wich param is default ?
+List of features that will be implemented (in no particular order) :
+
+- Integrate nalgebra cholesky decomposition update once the new version is released
+- Add [ndarray](https://docs.rs/ndarray/) support behind a feature flag
+- Add [serde](https://docs.rs/ndarray/) support behind a feature flag
+- Add simple [kernel regression](https://en.wikipedia.org/wiki/Kernel_regression#Nadaraya%E2%80%93Watson_kernel_regression) (not as clever but much faster).
+- add [benchmarks](http://www.resibots.eu/limbo/release-2.0/reg_benchmarks.html) to confirm correctness and validate performances
+- Add sparse gaussian proces
+- Improve test coverage
+
+List of API modifications that are being considered :
+
+- replace the builder pattern with a macro (might rely on [duang](https://crates.io/crates/duang))
+- simplify the kernel trait
+
+*Do not hesitate to start an issue if you want to contribute to one of those points.*