blosc2-src 0.1.4

FFI bindings for c-blosc2
Documentation
Here there are stored testing complementary files that are available for
all users to test their plugins.

The folder files are multidimensional arrays created by b2nd using
the 'example_frame_generator.c' example.
(https://github.com/Blosc/c-blosc2/blob/main/examples/b2nd/example_frame_generator.c).
Moreover, they have the next parameters:
- nthreads = 1;
- splitmode = BLOSC_ALWAYS_SPLIT;
- compcodec = BLOSC_ZSTD;
- complevel = 9;

Moreover, each dataset has different types and shapes:

**example_rand.b2nd**:

- ndim = 3;
- type = int;
- typesize = 4;
- shape[8] = {32, 18, 32};
- chunkshape[8] = {17, 16, 24};
- blockshape[8] = {8, 9, 8};

**example_same_cells.b2nd**:

- ndim = 2;
- type = int;
- typesize = 4;
- shape[8] = {128, 111};
- chunkshape[8] = {32, 11};
- blockshape[8] = {16, 7};

**example_some_matches.b2nd**:

- ndim = 2;
- type = long;
- typesize = 8;
- shape[8] = {128, 111};
- chunkshape[8] = {48, 32};
- blockshape[8] = {14, 18};

**example_float_cyclic.b2nd**:

- int8_t ndim = 3;
- type = float;
- typesize = 4;
- int64_t shape[] = {40, 60, 20};
- int32_t chunkshape[] = {20, 30, 16};
- int32_t blockshape[] = {11, 14, 7};

**example_double_same_cells.b2nd**:

- int8_t ndim = 2;
- type = double;
- typesize = 8;
- int64_t shape[] = {40, 60};
- int32_t chunkshape[] = {20, 30};
- int32_t blockshape[] = {16, 16};

**example_day_month_temp.b2nd**:

- int ndim = 2;
- type = float;
- typesize = 4;
- int64_t shape[] = {400, 3};
- int32_t chunkshape[] = {110, 3};
- int32_t blockshape[] = {57, 3};

This frame simulates values for: 
- Day: between 1 and 31 (column 0)
- Month: between 1 and 12 (column 1)
- Temperature: between -20 and 40 (column 2)

These fields are grouped by columns, so the frame has 400 rows that 
represent different days with a value for each of the three columns.

**example_item_prices.b2nd**:

- int ndim = 3;
- type = float;
- typesize = 4;
- int64_t shape[] = {12, 25, 250};
- int32_t chunkshape[] = {8, 10, 50};
- int32_t blockshape[] = {4, 5, 10};
    
This frame simulates item prices based on 3 dimensions:
- Month (dim 0): items are more expensive depending on the month
- Store ID (dim 1): items are more expensive depending on the store type
- Item ID (dim 2): each item has a different base price

Depending on each dimension index, the final price is calculated for 
each situation.