b'Data trends Data trendsof domains). It is suited to high density,(Smoothed Multivariate Mosaic). Be continuous data sets such as HyLoggerwarned that smoothing can move data and full suite downhole logs. boundaries and the example shows Figure 1 displays the result for athe result of maximum smoothing. downhole log of gamma, neutron count,Pseudo-logs of boundaries highlight resistivity and point resistance. On thethe downhole intervals with large signal left sit the four input signals overlyingvariation, shown by shades of green, for a y-axis. To the right are two mosaics ofthe gamma and neutron counts.domain boundaries using multivariateIn addition, the machine learning analysis and two pseudo-logs using theoptions can combine geology logs and univariate options. statistical clustering to assign rock types The default multivariate settingsto domains. A defined drill hole can be returned many small domains reflectingused to automatically assign rock types the high frequency signals (Defaultacross other holes in the area. Results Multivariate Mosaic). This was simplifiedare available as text files or images. Not Tim Keepingusing the Data Filter option to smooththe final answer, but a good start for a Associate Editor for geophysicalthe downhole logs before analysis andgeologist who needs to process many datamanagement and analysis only derive low frequency domainsdrill holes with numerical data.technical-standards@aseg.org.auData Mosaica practical application of the continuous wavelet transformA previous column explored the Continuous Wavelet Transform (CWT), and a perfect example of its application is available from CSIRO. June Hill (CSIRO) has applied the CWT to produce the Data Mosaic web application (https://datamosaic.geoanalytics.group/), an easy to use combination of signal analysis, machine learning statistics and visualisation to produce first pass stratigraphic and lithologic picks.The CWT is the shotgun of Fourier Transforms in that it searches for all possible wavelengths in data. Data Mosaic uses this to detect apparentFigure 1.Input signals, two mosaic and two pseudo-logs produced by Data Mosaic. Pseudo-log shades of boundaries in the signals and returngreen indicate magnitude variation of a chosen signal within a domain (light = low variation, dark = high plots of distinctive segments (a mosaicvariation).The ASEG in social mediaHave you liked/followed/subscribed to our social media channels? We regularly share relevant geoscience articles, events, opportunities and lots more. Subscribe to our Youtube channel for recorded webinars and other content. Email our Communications Chair Millicent Crowe at communications@aseg.org.au for suggestions for our social media channels.Facebook: https://www.facebook.com/AustralianSocietyOfExplorationGeophysicistsLinkedIn company page: https://www.linkedin.com/company/australian-society-of-exploration-geophysicists/Twitter: https://twitter.com/ASEG_newsYouTube: https://www.youtube.com/channel/UCNvsVEu1pVw_BdYOyi2avLgInstagram: https://www.instagram.com/aseg_news/ JUNE 2022 PREVIEW 40'