What are the differences between between zero knowledge succint non interactive of knowledge proof (zk snarks) and zero knowledge range proofs (zkrp) ?
A very good explanation from ING's paper on ZKRP:
ZK-SNARKS, are a ZKP-variant, that requires no interaction necessary between the prover and verifier. The most appealing example of ZK-SNARK is its ap- plication in Zcash . Also, Ethereums Metropolis will introduce ZK-SNARKS. This shows that ZK-SNARKS can, indeed, be applied in practice.
ZK-SNARKS are generic; they can verify any function which is particularly relevant for Ethereum which provides the Turing-complete EVM, allowing devel- opers to build any type of logic. Generic ZKPs like ZK-SNARKS, however come at a price. They are notably less efficient than specific zero-knowledge proofs.
Zero-Knowledge Range Proofs (ZKRP) allow for proofing that a number lies within a certain range.
Proofing that a number lies within a certain range is a more specific ZKP solution, which can be implemented more efficiently (-> gas usage) than the generic ZK-SNARKS.