Sparse Arrays 1461 92.1 Compressed Sparse Column (CSC) Sparse Matrix Storage . . . . . . . . . . . . . . . 1461 92.2 Sparse Vector Storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . comparable systems include: • Free and open source (MIT licensed) • User-defined types are as fast and compact as built-ins • No need to vectorize code for performance; devectorized code is fast • Designed if there are leading zero digits which don’t contribute to the value, they count for determining storage size of a literal. So 0x01 is a UInt8 while 0x0001 is a UInt16. That allows the user to control

Sparse Arrays 1461 92.1 Compressed Sparse Column (CSC) Sparse Matrix Storage . . . . . . . . . . . . . . . 1461 92.2 Sparse Vector Storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . comparable systems include: • Free and open source (MIT licensed) • User-defined types are as fast and compact as built-ins • No need to vectorize code for performance; devectorized code is fast • Designed if there are leading zero digits which don’t contribute to the value, they count for determining storage size of a literal. So 0x01 is a UInt8 while 0x0001 is a UInt16. That allows the user to control

Sparse Arrays 1708 94.1 Compressed Sparse Column (CSC) Sparse Matrix Storage . . . . . . . . . . . . . . . 1708 94.2 Sparse Vector Storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . comparable systems include: • Free and open source (MIT licensed) • User-defined types are as fast and compact as built-ins • No need to vectorize code for performance; devectorized code is fast • Designed if there are leading zero digits which don’t contribute to the value, they count for determining storage size of a literal. So 0x01 is a UInt8 while 0x0001 is a UInt16. That allows the user to control

Sparse Arrays 1708 94.1 Compressed Sparse Column (CSC) Sparse Matrix Storage . . . . . . . . . . . . . . . 1708 94.2 Sparse Vector Storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . comparable systems include: • Free and open source (MIT licensed) • User-defined types are as fast and compact as built-ins • No need to vectorize code for performance; devectorized code is fast • Designed if there are leading zero digits which don’t contribute to the value, they count for determining storage size of a literal. So 0x01 is a UInt8 while 0x0001 is a UInt16. That allows the user to control

Sparse Arrays 1708 94.1 Compressed Sparse Column (CSC) Sparse Matrix Storage . . . . . . . . . . . . . . . 1708 94.2 Sparse Vector Storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . comparable systems include: • Free and open source (MIT licensed) • User-defined types are as fast and compact as built-ins • No need to vectorize code for performance; devectorized code is fast • Designed if there are leading zero digits which don’t contribute to the value, they count for determining storage size of a literal. So 0x01 is a UInt8 while 0x0001 is a UInt16. That allows the user to control

Sparse Arrays 1773 96.1 Compressed Sparse Column (CSC) Sparse Matrix Storage . . . . . . . . . . . . . . . 1773 96.2 Sparse Vector Storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . comparable systems include: • Free and open source (MIT licensed) • User-defined types are as fast and compact as built-ins • No need to vectorize code for performance; devectorized code is fast • Designed if there are leading zero digits which don’t contribute to the value, they count for determining storage size of a literal. So 0x01 is a UInt8 while 0x0001 is a UInt16. That allows the user to control

Sparse Arrays 1771 96.1 Compressed Sparse Column (CSC) Sparse Matrix Storage . . . . . . . . . . . . . . . 1771 96.2 Sparse Vector Storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . comparable systems include: • Free and open source (MIT licensed) • User-defined types are as fast and compact as built-ins • No need to vectorize code for performance; devectorized code is fast • Designed if there are leading zero digits which don’t contribute to the value, they count for determining storage size of a literal. So 0x01 is a UInt8 while 0x0001 is a UInt16. That allows the user to control

Sparse Arrays 1770 96.1 Compressed Sparse Column (CSC) Sparse Matrix Storage . . . . . . . . . . . . . . . 1770 96.2 Sparse Vector Storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . comparable systems include: • Free and open source (MIT licensed) • User-defined types are as fast and compact as built-ins • No need to vectorize code for performance; devectorized code is fast • Designed if there are leading zero digits which don’t contribute to the value, they count for determining storage size of a literal. So 0x01 is a UInt8 while 0x0001 is a UInt16. That allows the user to control

Sparse Arrays 1770 96.1 Compressed Sparse Column (CSC) Sparse Matrix Storage . . . . . . . . . . . . . . . 1770 96.2 Sparse Vector Storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . comparable systems include: • Free and open source (MIT licensed) • User-defined types are as fast and compact as built-ins • No need to vectorize code for performance; devectorized code is fast • Designed if there are leading zero digits which don’t contribute to the value, they count for determining storage size of a literal. So 0x01 is a UInt8 while 0x0001 is a UInt16. That allows the user to control

Arrays 1761 95.1 Compressed Sparse Column (CSC) Sparse Matrix Storage . . . . . . . . . . . . . . . 1761CONTENTS xii 95.2 Sparse Vector Storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . comparable systems include: • Free and open source (MIT licensed) • User-defined types are as fast and compact as built-ins • No need to vectorize code for performance; devectorized code is fast • Designed if there are leading zero digits which don’t contribute to the value, they count for determining storage size of a literal. So 0x01 is a UInt8 while 0x0001 is a UInt16. That allows the user to control