This video shows a tutorial on how to navigate 2D NMR in basic mode.

In Basic mode, 

• "Fit" - Fits a cluster of peaks after selecting desired number of peaks

While REDEN automatically estimates the number of peaks, the user can easily adjust this value. 

The fitting process is completed in just a few seconds, and the result is immediately displayed to the user. 

The user can then examine the fitting by manipulating the cluster and viewing it from different angles via mouse movements. 

Results: 

• "Show Fit" - displays the overlay of the actual data and fitting model from the left panel

• "Show Fit Contour" - displays a contour plot of the cluster and fitting

• "Show Each Peak" - displays individual peaks

• "Show Residuals" - displays a residual plot of the fitting

• "Show Parameters" - displays the fitting parameters that provide peak amplitude, line-width, skewness, and peak center

If the fitting result is unsatisfactory, the user can repeat the process with a different number of peaks to fit. Once the optimal peaks have been picked by the decomposition, they can be effortlessly imported back into the spectrum by simply clicking the "Import to POKY" button.

This video shows a tutorial on how to navigate 2D NMR in advanced mode. It includes fine tuning tools that is not found in basic mode. These fine tuning tools are mainly for troubleshooting. 

REDEN's "Advanced" mode provides users with a wealth of data and extensive fine-tuning capabilities. This mode offers additional buttons for displaying intermediate processing steps, including 3D and contour plots, which can be useful in defining a cluster of peaks. Given the challenge of accurately identifying a cluster when peaks are closely spaced, these intermediate plots offer insights into the processing steps that led to the final result. If a mistake was made in selecting a nearby cluster, the user can adjust the viewing window and re-run REDEN as needed. as Aside from the peak display options in the “Basic” mode described above, the “Advanced” mode provides a way to display different aspects of the selected data region. 

When one of these buttons is pressed, REDEN performs the wavelet denoising in real time and use the cleaner data. Buttons with the “Contour” in the label show 2D contour plots while the buttons without it show 3D mesh plots. 

• 4 intermediate processing steps

  • Original (with contour) - show the fits on the original data that the user selects

  • Cluster (with contour) - shows the 3D plot of the isolated cluster based on image processing techniques

  • Denoise (with contour) - shows the fits on the denoised data

  • Zoom (with contour) - shows only the identified main cluster while everything else including other nearby clusters are hidden

REDEN uses default values for fitting in the “Basic” mode, but sometimes the cluster is so out of shape that the default values do not result in a good fit. In such cases, the user can adjust the "Amp Coef" (the coefficient of amplitude) or "L-W Coed" (the coefficient of linewidths) parameters to attempt another fit in the “Advanced” mode. Usually, just adjusting these parameters once will improve the fit of the cluster. 

REDEN provides four fitting algorithms: Gaussian, Lorentzian, a fast/optimized Lorentzian (called "F-Lorentzian"), and Pseudo-Voigt (Zaghloul and Ali, 2012). Users can switch between these options in the “Advanced” mode. 

• 4 fitting algorithm options

  • Gaussian 

    • Formulated with the SciPy’s multivariate_normal function

  • Lorentzian

    • Formulated with SciPy’s multivariate_t function

  • F-Lorentzian (Fast/Optimized Lorentzian) *default

    • Lorentzian function that we formulated the multivariate_t function with NumPy’s mathematics functions

  • Pseudo-Voigt

    • Combination of Lorentzian and Gaussian functions with the linear combination scale between them

• Voigt Scale – linear combination scale between Gaussian and Lorentzian

By default, the fast and highly optimized version of the Lorentzian algorithm is used, as also employed in Basic mode. Our tests indicate that this algorithm is twice as fast as the other available options. Again, users can change functions to obtain the most satisfactory results without hassles. 

This video shows a tutorial on how to navigate 3D NMR in basic mode.

The process for decomposition peaks in 3D NMR is identical to that for 2D NMR. 

• "Fit" - Fits a cluster of peaks after selecting desired number of peaks

While REDEN automatically estimates the number of peaks, the user can easily adjust this value. 

The fitting process is completed in just a few seconds, and the result is immediately displayed to the user. 

The user can then examine the fitting by manipulating the cluster and viewing it from different angles via mouse movements. 

Results: 

• "Show Fit" - displays the overlay of the actual data and fitting model from the left panel

• "Show Fit Contour" - displays a contour plot of the cluster and fitting

• "Show Each Peak" - displays individual peaks

• "Show Residuals" - displays a residual plot of the fitting

• "Show Parameters" - displays the fitting parameters that provide peak amplitude, line-width, skewness, and peak center

If the fitting result is unsatisfactory, the user can repeat the process with a different number of peaks to fit. Once the optimal peaks have been picked by the decomposition, they can be effortlessly imported back into the spectrum by simply clicking the "Import to POKY" button.

The main distinction is that 3D NMR produces 4D data because it includes additional peak intensity, so visualizing the data as a 3D plot will only display a single plane. To address this, the 3D REDEN module includes supplementary buttons that enable users to navigate the third axis of the data. 

This video shows a tutorial on how to navigate 3D NMR in advanced mode. It includes fine tuning tools that is not found in basic mode. These fine tuning tools are mainly for troubleshooting.

REDEN's "Advanced" mode provides users with a wealth of data and extensive fine-tuning capabilities. This mode offers additional buttons for displaying intermediate processing steps, including 3D and contour plots, which can be useful in defining a cluster of peaks. Given the challenge of accurately identifying a cluster when peaks are closely spaced, these intermediate plots offer insights into the processing steps that led to the final result. If a mistake was made in selecting a nearby cluster, the user can adjust the viewing window and re-run REDEN as needed. as Aside from the peak display options in the “Basic” mode described above, the “Advanced” mode provides a way to display different aspects of the selected data region. 

When one of these buttons is pressed, REDEN performs the wavelet denoising in real time and use the cleaner data. Buttons with the “Contour” in the label show 2D contour plots while the buttons without it show 3D mesh plots. 

• 4 intermediate processing steps

  • Original (with contour) - show the fits on the original data that the user selects

  • Cluster (with contour) - shows the 3D plot of the isolated cluster based on image processing techniques

  • Denoise (with contour) - shows the fits on the denoised data

  • Zoom (with contour) - shows only the identified main cluster while everything else including other nearby clusters are hidden

REDEN uses default values for fitting in the “Basic” mode, but sometimes the cluster is so out of shape that the default values do not result in a good fit. In such cases, the user can adjust the "Amp Coef" (the coefficient of amplitude) or "L-W Coed" (the coefficient of linewidths) parameters to attempt another fit in the “Advanced” mode. Usually, just adjusting these parameters once will improve the fit of the cluster. 

REDEN provides four fitting algorithms: Gaussian, Lorentzian, a fast/optimized Lorentzian (called "F-Lorentzian"), and Pseudo-Voigt (Zaghloul and Ali, 2012). Users can switch between these options in the “Advanced” mode. 

• 4 fitting algorithm options

  • Gaussian 

    • Formulated with the SciPy’s multivariate_normal function

  • Lorentzian

    • Formulated with SciPy’s multivariate_t function

  • F-Lorentzian (Fast/Optimized Lorentzian) *default

    • Lorentzian function that we formulated the multivariate_t function with NumPy’s mathematics functions

  • Pseudo-Voigt

    • Combination of Lorentzian and Gaussian functions with the linear combination scale between them

• Voigt Scale – linear combination scale between Gaussian and Lorentzian

By default, the fast and highly optimized version of the Lorentzian algorithm is used, as also employed in Basic mode. Our tests indicate that this algorithm is twice as fast as the other available options. Again, users can change functions to obtain the most satisfactory results without hassles. 

REDEN implements the axis order defined in the selected view window of POKY for the 3D spectrum analysis. 

• "Max z-planes" - defines how many planes through z-dimension will be used

If the cluster is not identified, not a whole, or overlapped with the other cluster, REDEN will advise the user what to do for identifying a cluster successfully. The user will need to zoom in, zoom out, or pan the spectral view extent if x- and y-dimensions need to be adjusted. If z-dimension needs to be adjusted, the user will use different value in the “Max z-planes” box.