float32 support

Author: vitalsong

README.md

@@ -50,7 +50,18 @@ arr_cmplx y6 = x1 * 2i;
 ### Slicing
 The behavior of slices is as close as possible to numpy. Except for cases with invalid indexes, in which case numpy does not throw an exception.
 ```cpp
-using namespace dsplib;
+arr_real x = {0, 1, 2, 3, 4, 5, 6};
+x.slice(0, 2) ///{0, 1}
+x.slice(2, -1) ///{2, 3, 4, 5}
+x.slice(-1, 0, -1) ///{6, 5, 4, 3, 2, 1}
+x.slice(-1, 0) ///OUT_OF_RANGE, but numpy returns []
+x.slice(0, -1, -1) ///OUT_OF_RANGE, but numpy returns []
+x.slice(-8, 7) ///OUT_OF_RANGE, but numpy returns [0 1 2 3 4 5 6]
+```
+
+### Fast Fourier Transform:
+The FFT/IFFT calculation table is cached on first run. To eliminate this behavior, you can use the fft_plan object.
+```cpp
 arr_real x = randn(512);
 arr_cmplx y = fft(x);
 ```
@@ -91,7 +102,7 @@ arr_real x2 = awgn(x1, 10);
 arr_real y = xcorr(x1, x2);
 ```
 
-Simple Spectrum Analyze (16-bit scale):
+### Simple Spectrum Analyze (16-bit scale):
 ```cpp
 int nfft = 1024;
 arr_real x = randn(nfft) * 1000;