Click here for more information

---

Table of Contents

• Introduction
• Development process
• Inputs and outputs of a knowledge discovery algorithm
• Recent updates
• Features
• Installing GDAL
• Maintenance
• Try your first geoAnalytics program
• License
• Documentation
• Getting Help
• Discussion and Development
• Contribution to geoAnalytics
• Tutorials
  • Conversion
  • Imputation
  • Clustering
  • Pattern Mining
• Real-World Case Studies

---

Introduction

geoAnalytics is an open-source Python-based Machine Learning library developed to discover various forms of useful information hidden in the raster data. The algorithms provided in this library cover a wide-spectrum of machine learning tasks, such as imputation, image fusion, clustering, classification, one class classification, and pattern mining. This library being platform independent can run any operating system.

---

Flow Chart of Developing Algorithms in geoAnalytics

---

Inputs and Outputs of an Algorithm in geoAnalytics

---

Recent Updates

• Version 2025.7.2: In this latest version, the following updates have been made:
  • Included new algorithms in patternMining, CorrelatedPatternMining, CoveragePatternMining, FaultTolerantFrequentPatternMining, FrequentPatternMining, LocalPeriodicPatternMining, PartialPeriodicFrequentPatternMining, PartialPeriodicPatternMining, PartialPeriodicPatternInMultipleTimeSeries, PeriodicCorrelatedPatternMining, RecurringPatternMining, RelativeFrequentPatternMining, and PeriodicFrequentPatternMining for identifying the patterns.
  • Included new algorithms in normalization.
  • Test cases are implemented using geoanalytics package.

Total number of algorithms: 50+

---

Features

• ✅ Tested to the best of our possibility
• 🔋 Highly optimized to our best effort, light-weight, and energy-efficient
• 👀 Proper code documentation
• 🍼 Sample examples of using various algorithms at ./tests folder
• 🤖 Works with AI libraries such as TensorFlow, PyTorch, and sklearn.
• ⚡️ Supports Cuda
• 🖥️ Operating System Independence
• 🔬 Knowledge discovery in static data and streams
• 🐎 Snappy
• 🐻 Ease of use

---

Installing GDAL Package

GDAL is an important toolkit in our library. It is for converting the raster data in any format into a human readable text or CSV format. We have present the methods to install this toolkit using Conda environment on a machine running Ubuntu operating system.

sudo apt-get update && sudo apt upgrade -y && sudo apt autoremove

sudo apt-get install -y cdo nco gdal-bin libgdal-dev


pip install --global-option=build_ext --global-option="-I/usr/include/gdal" GDAL==`gdal-config --version`


python -m pip install --upgrade pip setuptools wheel
python -m pip install --upgrade gdal 

If the above two commands have failed to install gdal, then execute the following commands:

conda install -c conda-forge libgdal
conda install -c conda-forge gdal
conda install tiledb=2.2
conda install poppler

Once the above commands were executed, check the version information by typing the following command on the terminal:

ogrinfo --version

Installation using Anaconda (geoanalytics package).

1. Install and set up Anaconda. URL: https://linuxize.com/post/how-to-install-anaconda-on-centos-7

2. Create a virtual environment using conda. E.g., coda create --name geoAnalytics

3. Enter into virtual environment. E.g., conda activate geoAnalytics

4. Install python. E.g., conda install python

5. Install pycharm from the website

6. Open Pycharm and using VCS download the latest copy of geoAnalytics from GitHub

7. In the pycharm, add geoAnalytics as the interpreter

8. Open the terminal in pycharm, and execute the following command

      pip install mplcursors matplotlib sklearn pandas
   

---

Maintenance

Installation

     pip install geoAnalytics

Upgradation

    pip install --upgrade geoAnalytics

Uninstallation

    pip uninstall geoAnalytics 

Information

    pip show geoAnalytics

---

Try your first geoAnalytics program

$ python

Install the Necessary Ubuntu Packages

$ !apt update

$ !apt install -y nco cdo gdal-bin

$ !which ncrename

Output: /usr/bin/ncrename

$ !which cdo

Output: /usr/bin/cdo

$ !which gdal_translate

Output: /usr/bin/gdal_translate

Install the geoAnalytics package

!pip install -U geoanalytics

Knowing the information of geoAnalytics package

pip show geoanalytics

Output: 
        Name: geoanalytics
        Version: 2025.6.10.3
        Summary: This software is being developed at the University of Aizu, Aizu-Wakamatsu, Fukushima, Japan
        Home-page: https://github.com/UdayLab/geoanalytics
        Author: 
        Author-email: Rage Uday Kiran <[email protected]>
        License: GPLv3
        Location: /usr/local/lib/python3.11/dist-packages
        Requires: deprecated, discord.py, fastparquet, matplotlib, mplcursors, networkx, numba, numpy, pandas, Pillow, plotly, psutil, psycopg2-binary, resource, scikit-learn, shapely, sphinx, sphinx-rtd-theme, tqdm, urllib3, validators
        Required-by: 

Download any dataset with extension of both .img and .lbl from the below link

Link --> https://data.darts.isas.jaxa.jp/pub/pds3/sln-l-mi-5-map-v3.0/

Example Direction : --> https://data.darts.isas.jaxa.jp/pub/pds3/sln-l-mi-5-map-v3.0/lon042/data/

Upload the .lbl file .img file

Operatioin 1: Raster to CSV

Step1: Import RasterToTSV package from geoAnalytics.conversion

from geoanalytics.conversion import Raster2CSV

Step2: Pass the lbl file as input and give desired outputFile name also specify the inputBand value as well as ouputBand value

converter = Raster2CSV.Raster2CSV(inputFile='MI_MAP_03_S16E035S17E036SC.lbl', outputFile='Moon.csv', startBand=1, endBand=9)

Step3: Convert the raster file into a CSV file

converter.run()

Output: 
        Processing: MI_MAP_03_S16E035S17E036SC.lbl
        Done. Output saved to: Moon.csv

Operation 2 : Clustering the Data

Step 1.1: Read the obtained CSV into a dataframe

import pandas as pd
df = pd.read_csv('Moon.csv', sep='\t')
df

x	y	1	2	3	4	5	6	7	8	9
1061317.265	-485173.607	1928	3446	3859	3913	4026	3999	4236	5245	6513
1061332.071	-485173.607	1924	3480	3876	3930	4059	3996	4243	5234	6518
1061346.877	-485173.607	1904	3476	3834	3923	4047	3992	4238	5222	6523
1061361.684	-485173.607	1874	3452	3801	3897	3959	3988	4228	5210	6518
1061376.490	-485173.607	1907	3464	3777	3868	3974	3983	4218	5198	6504
...	...	...	...	...	...	...	...	...	...	...
1091566.583	-515482.151	2090	3719	4007	3996	4117	4056	4310	5374	6633
1091581.390	-515482.151	2098	3734	4038	4020	4177	4083	4327	5389	6659
1091596.196	-515482.151	2114	3767	4046	4040	4213	4110	4332	5393	6685
1091611.002	-515482.151	2123	3813	4019	4056	4214	4136	4331	5397	6711
1091625.809	-515482.151	2125	3784	4011	4051	4184	4136	4329	5402	6737
4194304 rows × 11 columns

Step2: Import any clustering algorithm from geoanalytics.clustering

!pip install fuzzy-c-means
from geoanalytics.clustering import FuzzyCMeans
obj = FuzzyCMeans.FuzzyCMeans(dataframe=df)

Step3: Execute the run() method to cluster the data

labels, centers = obj.run(n_clusters=4)

Step4: Print the labels

labels

x	y	labels
1061317.265	-485173.607	2
1061332.071	-485173.607	2
1061346.877	-485173.607	2
1061361.684	-485173.607	2
1061376.490	-485173.607	2
...	...	...
1091566.583	-515482.151	2
1091581.390	-515482.151	0
1091596.196	-515482.151	0
1091611.002	-515482.151	0
1091625.809	-515482.151	0
4194304 rows × 3 columns

Step5: Print the centers

centers

Output: 
      array([[2098.51620472, 3762.45416178, 4099.00372289, 4146.23025903,
              4295.80971825, 4229.78337628, 4443.6879668 , 5471.85298977,
              6790.96720993],
             [2274.54605523, 4040.75198398, 4374.89922669, 4395.57058102,
              4538.50885801, 4454.45212911, 4677.95182781, 5739.96854105,
              7095.06428559],
             [1984.82196498, 3578.0340764 , 3908.79548852, 3961.66894715,
              4108.90409919, 4049.30456217, 4255.81099407, 5258.24055464,
              6549.13133772],
             [1896.71176664, 3421.42972113, 3741.56611478, 3802.88625245,
              3946.94784255, 3896.3217178 , 4093.18671711, 5066.07543066,
              6327.35005223]])

Step6: To know the run time and memory consumption of Algorithm, print the statistical values

obj.getRuntime()
obj.getMemoryRSS()
obj.getMemoryUSS()

Output: 
      Total Execution time of proposed Algorithm: 407.7022657394409 seconds
      Memory (RSS) of proposed Algorithm in KB: 1655512.0
      Memory (USS) of proposed Algorithm in KB: 1634900.0

Step7: Save the labels and centers

obj.save(outputFileLabels='FuzzyCMeansLabels.csv', outputFileCenters='FuzzyCMeansCenters.csv')

Output: 
      Labels saved to: FuzzyCMeansLabels.csv
      Cluster centers saved to: FuzzyCMeansCenters.csv

Operation3: CSV to Raster

Step1: Import CSV2Raster from geoanalytics.conversion

from geoanalytics.conversion import CSV2Raster as CSV2Raster

Step2: Pass the dataFrame and give desired outputFile name

process = CSV2Raster.CSV2Raster(dataframe=labels,outputFile='FuzzyCMeans.tiff')

Step3: Execute the run() method

process.run()

Output: 
      (0, '')
      (139, 'Segmentation fault (core dumped)')
      (0, '')
      (0, 'Input file size is 2048, 2048\n0...10...20...30...40...50...60...70...80...90...100 - done.')
      (0, '')

Operation 4: Visualization

Step1: Import TiffViewer from geoanalytics.visualization

pip install rasterio
from geoanalytics.visualization import TiffViewer

Step2: Input the RasterFile or pass it into the parameter

viewer = TiffViewer.TiffViewer(inputFile='FuzzyCMeans.tiff')

Step3: Display the image with desired scaling map

1. gray scale

viewer.run(cmap='gray', title='TIFF Image')

2. Jet Scale

viewer.run(cmap='jet', title='TIFF Image')

---

License

---

Documentation

The official documentation is hosted on geoAnalytics.

---

Getting Help

For any queries, the best place to go to is Github Issues Github Issues.

---

Discussion and Development

In our GitHub repository, the primary platform for discussing development-related matters is the university lab. We encourage our team members and contributors to utilize this platform for a wide range of discussions, including bug reports, feature requests, design decisions, and implementation details.

---

Contribution to geoAnalytics

We invite and encourage all community members to contribute, report bugs, fix bugs, enhance documentation, propose improvements, and share their creative ideas.

---

Tutorials

0. Conversion

Conversion
CSV To Raster
Raster To CSV

1. Imputation

Imputation
Backward Fill
Forward Fill
Interpolation
Mean Imputation
Median Imputation
Mode Imputation
Number Imputation
Soft Imputation

2. Clustering

Clustering
KMeans
KMeansPP
MEANshift
AffinityPropagation
Agglomerative
DBScan
FuzzyCMeans
Gaussianmixture
HDBScan
OpticsClustering
Spectral

3. Pattern Mining

Pattern Mining
Frequent Pattern Mining

Real World Case Studies

1. Lunar data analytics

Go to Top