Thomas Lecaé's repositories
LCLU_Frenchguiana
In order to map LCLU in french-Guyana, few scripts were developped or adapted to enable either to automaticaly map either to explore cloudless mosaic and even automaticaly detect floodings with Sentinel 1 SAR data.
auto_metadata
Create a pdf metadata file from pg tables
BNVD-s-app
A R Shiny app to extract data stored in a postgres server and use it in a data-visualisation based on leaflet and plotly
data
data for metabase
diversity_indexes
This code allows to compute 10+ indexes used in biodiversity metrics
example-cogugaison
A repo to store datasets and code to illustrate the use of R package "COGugaison" created by Antuki (many thanks) https://github.com/antuki/COGugaison
FlowmapBlue
Flow map visualization tool
fs_spatial_data_app
Behind this pompous name it's just a try to create an app that aims at valuing a open dataset that tell where emitting industries are and what do they emit
introjs-website
Intro.js website and documentation
ml_api
The code (and model) for a machine learning API that takes text as input and returns a note between 0 and 4 using a prediction model (SVM) trained using the code stored in my repo "SentimentAnalysis"
nb_cvi_comparison
This is a notebook used to compare differents data sources with the aims to feed a discuss about how to produce a coherent Landcover layer focused on the vineyard on the french metropolitan territory.
onthefly_lcover
A flask endpoint that returns histogram values from a raster landcover based on lat/lon values passed in the url
participativMapping
an attempt to create an interactive map (using mapbox, php, mysql) to allow a logged user to create and update (and of course store) geometries with a comment.
pg-metadata-to-latex
Create LateX metadata file from tables stored in Postgres schema
quattromap
compare up to 4 maps at once
scrapping_gesteau
jupyter notebook created to web-scrap (and archive) planning documents from https://www.gesteau.fr/rechercher/sage
sentimentAnalysis
A Jupyter Notebook - made during a training organised by French Statistics Institute (INSEE) - that uses Natural Language Processing (NLP) - after a bit of webscraping - to predict sentiment from a comment or whatever the text is.
walkAndPark
A streamlit/folium app to access results of a processing that aimed to find the closest park for each feature of dataset and return the walking time to get to it
walkAndPark_backend
Code to compute the time to walk from a polygon dataset and their corresponding parcs