I'm looking to convert testint data to FRF graphs. How do I go from the arrays of data that I have to these FRFs?

Question

I'm a test engineer and I recently did a modal analysis test for one of our products. I have data streams for an impact hammer and several tri-axial accelerometers. The guys I work with are asking me to provide them FRF graphs of the data, which I've never done before. I'm currently trying to figure out how to use Scipy for this endeavor but failing. I used the signal.welch() function recently for a PSD analysis and it was very simple, all I had to do was pass one of my data arrays(e.g. Accel_x_1), the sampling rate (fs), and what kind of scaling I wanted. However, when I look at signal.freqresp() or signal.freqz() it requires a dlti from me, which I cannot figure out how to create.

    datapoints = [x1data, x2data, y1data, y2data, z1data, z2data]
    # MAKES THE PSD VALUES FOR GRAPHS
    plt.ion()
    for x in datapoints:
        psd_values = signal.welch(x, fs=2048, scaling='density')
        x_val = psd_values[0]
        y_val = psd_values[1]
        plt.plot(x_val, y_val)

The above is what I did for the PSD via signal.welch. Unfortunately, signal.freqresp and the other functions do not accept just an array of data. I'm not even sure if those are the .signal functions I need to use to reach my goal. I want to obtain FRF graphs for the amplitude, phase, and coherence.