I want to realize IFFT function by CUDA. Input data is complex. The real numbers are imported from phase_init_befroe_R.csv, and the imaginary numbers are imported from phase_init_before_C.csv. Then I run the cufftExecC2C or the cufftExecZ2Z function. The result is saved in out_R.csv and out_C.csv. out_R.csv is what I want.
When the macro SINGLE is defined, all the data will be single-precision. Here is the result drawn by Matlab command imagesc(csvread("out_R.csv"):

I want to realize IFFT function by CUDA. Input data is complex. The real numbers are imported from phase_init_befroe_R.csv, and the imaginary numbers are imported from phase_init_before_C.csv. Then I run the cufftExecC2C or the cufftExecZ2Z function. The result is saved in out_R.csv and out_C.csv. out_R.csv is what I want.
When the macro SINGLE is defined, all the data will be single-precision. Here is the result drawn by Matlab command imagesc(csvread("out_R.csv"):

I want to realize IFFT function by CUDA. Input data is complex. The real numbers are imported from phase_init_befroe_R.csv, and the imaginary numbers are imported from phase_init_before_C.csv. Then I run the cufftExecC2C or the cufftExecZ2Z function. The result is saved in out_R.csv and out_C.csv. out_R.csv is what I want.
When the macro SINGLE is defined, all the data will be single-precision. Here is the result drawn by Matlab command imagesc(csvread("out_R.csv"):

I want to realize IFFT function by CUDA. Input data is complex. The real numbers are imported from phase_init_befroe_R.csv, and the imaginary numbers are imported from phase_init_before_C.csv. Then I run the cufftExecC2C or the cufftExecZ2Z function. The result is saved in out_R.csv and out_C.csv. out_R.csv is what I want.
When the macro SINGLE is defined, all the data will be single-precision. Here is the result drawn by Matlab command imagesc(csvread("out_R.csv"):

I want to realize IFFT function by CUDA. Input data is complex. The real numbers are imported from phase_init_befroe_R.csv, and the imaginary numbers are imported from phase_init_before_C.csv. Then I run the cufftExecC2C or the cufftExecZ2Z function. The result is saved in out_R.csv and out_C.csv. out_R.csv is what I want.
When the macro SINGLE is defined, all the data will be single-precision. Here is the result drawn by Matlab command imagesc(csvread("out_R.csv"):

I want to realize IFFT function by CUDA. Input data is complex. The real numbers are imported from phase_init_befroe_R.csv, and the imaginary numbers are imported from phase_init_before_C.csv. Then I run the cufftExecC2C or the cufftExecZ2Z function. The result is saved in out_R.csv and out_C.csv. out_R.csv is what I want.
When the macro SINGLE is defined, all the data will be single-precision. Here is the result drawn by Matlab command imagesc(csvread("out_R.csv"):

I want to realize IFFT function by CUDA. Input data is complex. The real numbers are imported from phase_init_befroe_R.csv, and the imaginary numbers are imported from phase_init_before_C.csv. Then I run the cufftExecC2C or the cufftExecZ2Z function. The result is saved in out_R.csv and out_C.csv. out_R.csv is what I want.
When the macro SINGLE is defined, all the data will be single-precision. Here is the result drawn by Matlab command imagesc(csvread("out_R.csv"):