Release

Notes
Isatis.neo 2024.04
© Copyright 2024 Datamine / Geovariances Software
All Rights Reserved Confidential and Proprietary
Published: Wednesday, May 22, 2024
Legal Disclaimer
The product described in this documentation may be connected to, and/or communicate
information and data via, a network interface, which should be connected to a secure
network. It is your sole responsibility to ensure a secure connection to the network and to
establish and maintain appropriate measures (such as but not limited to the installation of
firewalls, application of authentication measures, encryption of data, installation of
antivirus programs, etc.) to protect the product, the network, your systems, and the
interface against any kind of security breach, unauthorised access, interference,
intrusion, leakage, damage, or corruption or theft of data. We are not liable for damages
or losses related to any such security breach, unauthorised access, interference,
intrusion, leakage, damage, or corruption or theft of data.
 3

Contents
Overview 4

Licensing 5
Project compatibility 5
Platforms and Requirements 5
Further Information 5

Isatis.neo 2024.04 6

General Improvements 6
Python 3.11 6
Move drag/drop 7
Batch and Python tutorial 7
Home 8
Import Fusion 8
Import Isatis.neo 10
Export NetCDF 12
Statistics 14

EDA - Variogram lag slider 14

Quantiles calculation 15
PCA/MAF backward transformation 16
Interpolation 17
Accuracy plot 17
Cross-Validation - Output selection 19
Early Access 19
Grade post-processing 19
Defect Fixes 21
Overview 4

Overview
Isatis.neo is a powerful, intuitive, and highly customizable software solution in
geostatistics. It provides geologists and resource teams with an unrivaled range of
geostatistical tools, machine learning, and Python coding to enable precise resource
modeling and confident decision-making. Isatis.neo allows in-depth data analysis and
visualization, complex geological and geometallurgical analysis, and in-situ and
recoverable resource estimation. The software also enables extensive uncertainty and
risk analyses based on simulations.
Overview 5

Licensing
Before installation of the new Isatis.neo version, please check that your license is still
covered by a maintenance contract and/or your license key is still available. By default,
license keys are valid for 12 months. They are compatible with all the software versions
that will be released during the validity period of your maintenance contract.
Site licenses need to be installed on your license server by your license administrators.
For the cloud licenses, a new license file will be automatically deployed on Geovariances’
servers.
Starting with Isatis.neo-mining 2023.08.1, we provide a special version (on Datamine’s
Freshdesk website) compatible with Datamine License services as well as a version that
works with Geovariances License Manager (RLM, from Geovariances’ website). Please
ensure you use the version compatible with your license.
Please contact us at [email protected] for any information regarding your
license and maintenance contract.

Project compatibility
Your projects are automatically converted into the new format when you open them.

Platforms and Requirements

Isatis.neo is available on PC/Windows 10 or 11 (64 bits).

Further Information
This document includes cumulative release notes for Isatis.neo. Release notes for other
versions of Isatis.neo are available via the Support Portal or via the Geovariances’
website.
Isatis.neo 2024.04 6

Isatis.neo 2024.04
General Improvements
Python 3.11
The Python version embedded in Isatis.neo has been upgraded from version 3.7 to
version 3.11.
This allows you to use more recent versions of the secondary Python modules and benefit
from the latest functionality of those packages.
The main Python modules that have been updated to their latest versions are the
following: NumPy, Scipy, Pandas, and Scikit Learn. Other secondary packages are also
updated (Matplotlib, PySide2,..)

Python scripts migration advice

There are no major changes in the Python syntax between versions 3.7 and 3.11. Most of
your scripts should only require minor modifications.
Those minor modifications should be mostly some methods / classes renamed in the 3rd
party Python packages mentioned above. If you encounter deprecated functionality in
those modules, please refer to the changes in those Python packages.
For users that embed additional external Python modules, in Batch files for example,
extra attention must be paid to them.
Their version might also need to be updated if they are not compatible with the more
recent base Python modules that are installed with Isatis.neo.
If the package of the concerned module was specific to Python 3.7, you will also need to
replace it with the version of that package that is distributed for Python 3.11.

Impact on Vulcan imports/exports

The tools in Isatis.neo to access Maptek Vulcan data rely on a Python module provided by
Vulcan. Up to Vulcan 2023.3, the needed Python version was 3.7. From Vulcan 2023.4,
that Maptek module needs Python version 3.11. This means that starting from Isatis.neo
2024.04, The Vulcan tools will require at least Maptek Vulcan 2023.4 installed (multiple
Vulcan versions can be installed to keep compatibility with older Isatis.neo versions and
the new version).
Isatis.neo 2024.04 7

Move drag/drop
It is now possible to organize your Data tree with drag & drop. You can move several
files/folders into another folder more easily than before (you had to use the task Move
Data to move each file individually).

Note: The drag & drop can also be recorded in a Batch file.

Batch and Python tutorial

A new tutorial is available in the Tutorials section of the Project manager. It aims at
showing with examples how to interface Python and Isatis.neo through the use of the
Calculator and the Batch. It makes use of some Python packages installed by default in
Isatis.neo like matplotlib, pandas. The Batch logic and the production of graphics are
also illustrated in this tutorial.
Isatis.neo 2024.04 8

Home
Import Fusion
A new import task is available to import drillholes from Fusion databases. This import is
reachable from the Home section in the ribbon, under the Datamine (*.dm) import:

Note: This task is very similar to the ODBC import task.

Isatis.neo 2024.04 9

To import your drillholes, you have to connect to the server where the database is stored.
You have 2 ways to do that:
l Open the Fusion Exporter assistant: Having Fusion on the same machine is not
required here because when installing Isatis.neo, the Fusion Exporter Assistant is
installed too. It is an interface provided by Fusion to access your databases easily.

l Load a previous import parameters file: You can select a previously generated
.json file containing the parameters needed to connect your Fusion database.
Then, you have to define the roles of your data tables:
Isatis.neo 2024.04 10

l The Tops/Collars table which corresponds to the Collar is mandatory.

l The Deviations table which corresponds to the Survey.
l The Core samples/Assay, Logs, and Markers tables can be added by pressing the
Add button.
You can choose one of the following 4 desurveying methods: Spherical Arcs,
Interpolated Azimuth and Dip, Tangential, Balanced Tangential.
Once all the parameters are filled, you can choose each role/unit class… for each
variable of each data table (as in the different drillholes import tasks). A summary of your
import is displayed at the end of the task.

Import Isatis.neo
Another new import task is available to import different objects between Isatis.neo
projects. This import is reachable from the Home section in the ribbon:

You select the input project containing the data you want to import: it must come from
compatible projects: same coordinate system, same version, same application
(standard, mining, petroleum), same database version, etc.
Isatis.neo 2024.04 11

You can import several types of data at the same time (available in the multiple tabs
located on the left part of the task):
l Datasets (2D/3D points, grid, drillholes, meshes…):
l If the dataset selected already exists in your project (grid/data table…), the
variables selected will be copied into your dataset, only if the geometry
matches.
If the geometry doesn’t match, the name of the dataset will be renamed (ex:
XXXX (2)).
l If the variable already exists in the dataset, the name will be replaced (ex:
XXXX (2)).
l etc.
l Geostatistical sets, Chart files (graphics and tables), Scenes (2D/3D), Color
scales, Neighborhoods, Files in the public folder, Batch files
The objects will be copied, and if one or several objects with the same name already
exist, the names will be changed (ex: XXXX (2)).
For the Scenes, if a section exists and is linked to the scene, the section will be
imported too.
A summary of the imported data is displayed in the Messages window.
Isatis.neo 2024.04 12

If any data has been renamed, the table line is colored orange.

Export NetCDF
A new export task has been created to export NetCDF files as Grid or Points (2D or 3D).

There are 2 modes available: Export as grid and Export as points.

l Export as grid: only regular grid datasets can be selected to run the export.

Note: If the grid is rotated, you cannot export Text or Categorical variables.
l Export as points: all the data tables can be selected to run the export.
Isatis.neo 2024.04 13

A Selection can be applied to the data table to only export a part of your dataset. You can
Export All Variables, or only some of them. You can also choose the coordinate system
in which you wish your data to be stored. A projection header file (.prj) can be stored with
the output file. If the chosen coordinate system is different from the original one or if the
grid is rotated, a re-interpolation will be applied.

Note: It is not possible to write netCDF files that exceed 2 GB.

Isatis.neo 2024.04 14

Statistics
EDA - Variogram lag slider
In the 2023.12 version, a new interactive feature was added to the Exploratory Data
Analysis task for the computation of variograms: the lag slider. The Slider can be
activated by clicking on the button. With the Slider enabled, it will pre-compute all the
variograms for all multiple lags of the Slider step value up to a maximum (= Maximum
distance / 2).
The pre-computation of the variograms was extremely long (mainly with omnidirectional
variograms) so we improved the performance by changing several computation
parameters. When the slider is activated:
l The lag tolerance is now fixed to 50%
l No refined lag can be computed.

The calculation time is now much better than before: with or without the slider activated,
the task takes the same amount of time.
Isatis.neo 2024.04 15

We also fixed several issues due to the new slider mode:

l In the original version (2023.12), the interactive slider management broke some
update mechanism in EDA. It was especially visible when the Auto apply was off:
modifying any parameter related to the slider (i.e. granularity, lag or maximum
distance) led to a recalculation update while it should not.
l Isatis.neo asks the user to define a lag value and a maximum distance. But
internally, to copy the behaviour of Isatis Classic, the latter parameter used to be
converted to a number of lags. A new maximum distance would be calculated,
replacing the user-entered value.
l No mechanism existed to globally set the vertical bounds for the different slider-
step-lag values leading to many changes in the vertical axis bounds when using the
interactive slider. So we fixed it to the 90% quantile (in order to filter possible high
values).

Quantiles calculation
Isatis.neo previously calculated quantiles by first sorting the data. Sometimes this sorting
is necessary (when computing NP-plots), but when only a fixed set of quantiles is needed,
another algorithm to find the wanted quantile(s) will usually be faster. So we implemented
Andrei Alexandrescu’s adaptive quickselect to improve the performances when
computing the quantiles.
Isatis.neo 2024.04 16

This improvement is visible in the Statistics and Statistics by Category tasks. With large
datasets, quantile calculation was sometimes not possible (Not enough memory) and
sometimes very slow.
Here is some comparative tests between 2 versions of Isatis.neo (2023.12.1 and
2024.04):

PCA/MAF backward transformation

It is now possible to generate the backward transformation of macro variables as
simulations. The index All indices can be selected in each row. In this case, a macro
variable (one index per realization) will be automatically generated on output.
Isatis.neo 2024.04 17

Interpolation
Accuracy plot
A new tool is available to check the model of local uncertainty: the Accuracy plot. It
compares the probability interval against the actual fraction of true values that are inside
the probability interval.
Example: For a probability interval of 50% (defined by the P25 and P75), we expect that
over the study area, 50% of the true values are inside this probability interval.
The procedure is repeated for a series of probability intervals to create the accuracy plot.
To create the plot, the task needs several parameters:
Isatis.neo 2024.04 18

l The input geostatistical set which contains an anamorphosis (gaussian variable).

You can select an optional Selection variable and modify the Variable of interest (if
several Gaussian variables are available in the Geoset)
l A Neighborhood (unique or moving).

Note: With a moving neighborhood, the parameters of the 2 tabs Capping and
Nested are disabled in this task.
l The Computation Mode. Two modes are available:
l Simple Kriging: This mode uses the cross-validation calculator to cross-
validate the Gaussian variables and create the standardized error.
l Turning Bands Simulations: Non-conditional simulations are performed on
input data and for each simulation, we cross-validate the variable and create
the estimated residual variable r* on input data. For this computation mode,
several parameters are mandatory:
l The number of realizations to be performed.
l The number of bands (used in the simulations task).
l The seed used in the TBS computation.
The graphic can be saved in a Chart File by clicking on the Store chart file button
available at the bottom left of the interface.
Isatis.neo 2024.04 19

Cross-Validation - Output selection

In this new version, we have added a new item to the Special option section: Use
customized output selection. It is now possible to calculate cross-validation on a sub-
part of the data by using either a selection variable or a categorical variable.

Early Access
Grade post-processing
The new task Grade Post-processing is located in the Early Access tab. It is designed to
downscale the resolution of your block model. This tool calculates local grade tonnage
curves on a panel grid (variables for the quantity of metal Q, tonnage T, and mean grade
above cutoff M) from estimated or simulated grade variables on a block model.
To generate the output variables, the task needs several parameters:
l The input data table which contains the estimated / simulated grades:
l You can select a Selection and / or a Weight variable.
l The Main variable and the Secondary variables can be either a Macro or a
Simple variable (simulations or estimation).
l You have to choose on which cutoffs the computation will be performed and how to
replace your N/A values (skipping the samples or replacing by zero or by the mean
value).
l The output grid can be either created or selected from existing ones.
Isatis.neo 2024.04 20

You may then specify several output variables:

l The QTM tab computes the Q, T, M for each cutoff, over all realizations and
stores all these Q, T, M. The output is a macro variable for each grade variable
defined on input (main and secondary variables) and for each cutoff. Each
output variable has one index for each realization.
The Coarsen mean, Proportions and Count variables are simple variables.
They are of particular interest to be stored in case of non-matching grids.
l The Statistics tab is only available if the Main variable is a macro variable. It
computes the Q, T, M for each realization, one by one, and stores only the
different statistics of these Q, T, M. The output is a macro variable for each
grade variable defined on input (main and secondary variables). Each output
variable has one index for each cutoff.
Tick the statistics you want to compute: Mean, Standard deviation, Variance,
Minimum, Maximum and/or Quantiles.
Isatis.neo 2024.04 21

Defect Fixes
ING-4256 - Simulations: The accumulation above threshold calculation was wrong. The
results are now transformed using the volume and density (or only the surface in 2D) like
it is done in the kriging.
ING-4362 - Modify 2D/3D: It was possible to change the dimension of several files but in
all cases, the compound variables which rely on those dimensions were not updated.
Now, the compound variables are also modified when changing dimensions.
ING-4505 - Variography: The variogram map did not change when you set an input
weight variable.
ING-4654 - Unfolding: Increasing the number of pillars for bend unfolding could lead to
very sparse unfolding models and so neglect much data to unfold. The algorithm has
been changed, possibly changing results.
ING-4736 - Exploratory Data Analysis: EDA wrongly let the user create Gaussian or
residuals transformation on a PPMT variable, leading to a crash when creating a
variogram item on it.
ING-4814 - Unfolding: Lines Unfolding could fail and pop an error message about an
existing Sel_count variable when relaunching unfolding.
ING-4875 - EDA / Variography: The experimental variogram was recalculated after the
first modification of the variogram model in multidirectional mode.
ING-4880 - Import Datamine: 'DM "Ellipsoid" files (with fields XP, YP, ZP, AZI, DIP,
ROLL, RAD1, RAD2, RAD3) can now be imported as points files.
ING-4943 - LAS Import: When importing LAS files in the mode "customized tops coordin-
ates", it was not possible to run a single las import in a loop to import multiple files.

Note: The recommended simpler way of doing this is to do a single LAS import with all
the LAS files directly, so the top coordinates can be defined in the UI instead of having
to write a script.

ING-4958 - Cross-Validation:A parameter has been created to generate output variables

for selected variables only. It is now possible to save only output variables linked to a
subset of the input variable.
ING-4963 - Isatis Migration: When trying to restore an Isatis classic backup, an error
might indicate that only projects greater than 16.0.0 can be upgraded even if it is. This
would generally happen if the customer never installed Isatis on this machine. The error
does not appear anymore and we can now restore an Isatis backup even if we never
installed Isatis.
Isatis.neo 2024.04 22

ING-4965 - Simulation Reduction: There was an error when the charts were empty so
we fixed it and we add an status error when the number of simulations or combinations
are too low for the genetic algorithm.
ING-4966 - EDA: The display of pairs in experimental variogram as histogram or curve
was broken when using the lag slider.
ING-4970 - Uniform Conditioning: We have set the tonnage correction maximum to 10
and a status error is displayed if the value is 0.
 Datamine enables efficient and sustainable
mining through the application of
world-leading technology and services.

Read the Docs

docs.dataminesoftware.com

Get in Touch
[email protected]
www.dataminesoftware.com/support

Find Us
AUSTRALIA | BRAZIL | CANADA | CHILE | CHINA |
ECUADOR | FRANCE | GHANA | INDIA | INDONESIA |
KAZAKHSTAN | MALAYSIA | MEXICO | MONGOLIA | PERU |
PHILIPPINES | SOUTH AFRICA | TURKEY | UNITED
www.dataminesoftware.com KINGDOM | USA | UZBEKISTAN