Analytics Page
The Analytics Page is for the extraction of useful information from a set of processed spectra using various methods, including average and standard deviation analysis, correlation heatmapping, derivative analysis, peak identification, hierarchical clustering, principal component analysis (PCA), and t-distributed stochastic neighbor embedding (t-SNE). These functions can be used to identify patterns and trends in the data, filter data for machine learning, visualize important parts of the data, and much more.
General Operation
To analyze a particular set of spectra, first upload and proccess the data. Then, in the processing page, ensure that the spectra of interest are listed under the section labeled “Select Spectra you want to visualize.” These are the spectra that will be used for analysis. Click the “x” next to unwanted spectra to remove them from the list.
Once the data of interest have been selected, go to the Analytics page and select the desired type of analysis using the drop-down menu. For further details, see the in-depth documentation for each of the analysis functions.