The setting of dt leads to different results?

[ZiyeLiao]

I have been analyzing daily samples over a span of 9 years. I found that when I set the parameter dt to 1/365, two globally significant main periods (2.378d, 0.992y) are detected, whereas setting dt to 1 results in no significant periods being detected. I also try dt=1/7, and only the period of 2.3d is sinificant. I am uncertain if I have set the parameter correctly. Which one is the most suitable?
Below is the code snippet I have been using.:

n = 1
freq = 'D'
t0 = 1
dt = 1/365 
s0 = -1
dj = 1/12
J = -1

def resample2signal(data, timestamp='time', freq='D', t0=2011, dt=1/365):
    series = data.resample(rule=freq, on=timestamp).size()
    series.name = 'counts'
    calls = series.values
    n = len(calls)
    time = np.arange(0, n) * dt + t0
    return time, calls

time, calls = resample2signal(dat, freq=freq, t0=t0, dt=dt)

p = np.polyfit(time - t0, calls, 1)
calls_notrend = calls - np.polyval(p, time - t0)
avg = calls_notrend.mean()
std = calls_notrend.std()
var = std ** 2
calls_norm = (calls_notrend - avg) / std

N = calls.size
mother = pycwt.Morlet(6)
wave, scales, freqs, coi, fft, fftfreqs = pycwt.cwt(calls_norm, dt, dj, s0, J, mother)

power = (np.abs(wave)) ** 2 
period = 1 / freqs 
power /= scales[:, None]

alpha, _, _ = pycwt.ar1(calls)  
signif, _ = pycwt.significance(1.0, dt, scales, 0, alpha,
                                        significance_level=0.95,
                                        wavelet=mother)
sig951 = np.ones([1, N]) * signif[:, None]
sig95 = power / sig951


glbl_power = power.mean(axis=1)
dof = N - scales
glbl_signif, tmp = pycwt.significance(var, dt, scales, 1, alpha,
                                    significance_level=0.95, dof=dof,
                                    wavelet=mother)

Thanks!