Skip to content
Snippets Groups Projects

#194: Add documentation for GraphView

Open Ghost User requested to merge hrouis/aidge_core:aidge_core#194 into dev
1 file
+ 14
12
Compare changes
  • Side-by-side
  • Inline
@@ -129,7 +129,7 @@ public:
* @brief Save the GraphView as a Mermaid graph in a .md file at the
* specified location.
* @param path: Path where the file should be put
* @param verbose @todo i have no idea what it's doing
* @param verbose If true give more informations in the console during the saving process
* @param showProducers if true, shows additional informations
*/
void save(const std::string& path, bool verbose = false, bool showProducers = true) const;
@@ -239,7 +239,7 @@ public:
*/
inline auto dataInputs(const std::string name) const { return mNodeRegistry.at(name)->dataInputs(); }
/** @todo the nullptr behavior is strange: how is a node defined as an "input node" of the graph?
/**
* @brief List outside input connections of the GraphView. The vector
* size is guaranteed to match the number of outside inputs of the GraphView. If there is
* no external connection to a given input, a pair of nullptr and gk_IODefaultIndex is returned.
@@ -247,7 +247,7 @@ public:
*/
std::vector<std::pair<NodePtr, IOIndex_t>> inputs() const;
/** @todo having two inputs function, one without args that treat graph's inpput and one with args treating a node's input is fishy
/**
* @TODO @warning what if the node isn't found? where is the try catch of the .at?
* @brief List all input connections (within and outside) of the specified GraphView node named "name".
* @return std::vector<std::pair<NodePtr, IOIndex_t>>
@@ -281,8 +281,8 @@ public:
* compatible with the selected kernel.
* If not, add a Transpose Operator.
* 4 - Propagate Tensor dimensions through the consecutive Operators.
@params string: backend @todo: explain params
@params Aidge Datatype: datatype
@params string: backend Backend used, default is cpu
@params Aidge Datatype: datatype used, default is float32
@params vector of vector of DimSize_t: dims
*/
void compile(const std::string& backend = "cpu",
@@ -404,8 +404,8 @@ public:
void setInputId(IOIndex_t inID, IOIndex_t newNodeOutID);
/**
//@todo here i am
* @brief Include a Node to the current GraphView object.
* @brief Include a Node to the current GraphView's set of Nodes.
* If the Node is named, it is added in the registery as well
* @details If the GraphView has no root Node (most likely this GraphView is empty)
* the added Node becomes the GraphView's root Node
* @param otherNode Node to add.
@@ -416,8 +416,9 @@ public:
/**
* @brief Include a set of Nodes to the current GraphView object.
* @param otherNodes
* @param includeLearnableParam
* @param otherNodes Nodes to add to the GraphView
* @param includeLearnableParam Include non-data inputs, like weights and biases
* in the GraphView automatically. Default: true.
* @return true if graph ordering is unique (meaning inputs/outputs order is well defined).
*/
bool add(std::set<NodePtr> otherNodes,
@@ -425,7 +426,7 @@ public:
/**
* @brief Include a set of Nodes to the current GraphView object.
* The first element of the otherNodes pair is the start node and
* The first element of the otherNodes pair is the new RootNode of the GraphView and
* the second is the remaining nodes to add.
* @param otherNodes
* @param includeLearnableParam
@@ -435,8 +436,8 @@ public:
bool includeLearnableParam = true);
/**
* @brief Include every Node inside another GraphView to the current
* GraphView.
* @brief Include every Node inside another GraphView to the current GraphView.
* @details If the current GraphView has no RootNode it takes the RootNode of the new GraphView
* @param other_graph GraphView containing the Nodes to include.
* @return true if graph ordering is unique (meaning inputs/outputs order is well defined).
*/
@@ -444,6 +445,7 @@ public:
bool includeLearnableParam = true);
/**
//@todo here i am
* @brief Include a Node in the current GraphView and link it to another
* already contained Node.
*
Loading