from rtlsdr import RtlSdr
import numpy as np
import matplotlib.pyplot as plt
import sys

if __name__ == "__main__":
	graph = 'n'
	center_frequency = 434.7e6
	if (len(sys.argv)) > 0:
#		print("There are arguments!")
		center_frequency = float(sys.argv[1]) - 200e3
		if (center_frequency == 0):
			center_frequency = 434.7e6
		if (len(sys.argv)) > 1:
#			print("There are more arguments")
			if (sys.argv[2] == 'g') or (sys.argv[2] == '-g'):
				graph = 'y'  
        
	sampling_rate = 1024e3 # 250e3 # Hz
	duration = 65536/sampling_rate # 1          # seconds
	
	try:	
		sdr = RtlSdr()
		
		# configure device
		sdr.sample_rate = sampling_rate # 250e3 # 2.4e6
		#center_frequency = 434.8e6
		sdr.center_freq = center_frequency
		sdr.gain = 47
		sdr.direct_sampling = False
		
		# signal = sdr.read_samples(64*1024) #256
		signal = sdr.read_samples(duration*sampling_rate).real #256
		
	#	print(f"Center frequency is {center_frequency}")
		
		sdr.close()
		
		# Compute the FFT
		fft_result = np.fft.fft(signal)
		
		# Calculate the frequencies corresponding to the FFT output
		n = len(signal)
		frequencies = np.fft.fftfreq(n, d=1/sampling_rate)
		
		# Take the absolute value for amplitude spectrum and consider only the positive frequencies
		positive_frequencies_indices = np.where(frequencies >= 0)
		positive_frequencies = frequencies[positive_frequencies_indices]
		amplitude_spectrum = 2/n * np.abs(fft_result[positive_frequencies_indices]) # Normalize for amplitude
		
		if (graph == 'y'):
		  # Plotting the results
		  t = np.linspace(0, duration, int(sampling_rate * duration), endpoint=False)
			
		  plt.figure(figsize=(12, 6))
		  
		  plt.subplot(1, 2, 1)
		  plt.plot(t, signal)
		  plt.title('Time Domain Signal')
		  plt.xlabel('Time (s)')
		  plt.ylabel('Amplitude')
		  
		  plt.subplot(1, 2, 2)
		  plt.stem(positive_frequencies, amplitude_spectrum, markerfmt=" ", basefmt="-b")
		  plt.title('Frequency Domain (FFT)')
		  plt.xlabel('Frequency (Hz)')
		  plt.ylabel('Amplitude')
		  plt.grid(True)
		  
		  plt.tight_layout()
		  plt.show()
		
		# print(amplitude_spectrum)
		x = amplitude_spectrum 
		# print(x)
		min_value = min(x)
		max_value = max(x) * 100
		
		#freq_min = np.argmax(min_value)
		# print(np.argmax(x))
		# print(np.argmax(x)*(150e3 - 10e3)/(9770 - 709))
		# print(sampling_rate)
		# print(center_frequency)
		
		offset = (np.argmax(x)*(150e3 - 10e3)/(9770 - 709))
		freq_max = center_frequency + offset + 2000
	except: # if no RTL-SDR or in use, stop scanning with high max value and center frequency
		max_value = 100
		freq_max = center_frequency + 200e3 # should be 434900000
		
	print(f" {freq_max:.0f} {max_value:.0f}")