README.md

PyVino

Projet d'identification et caracterisation des vins avec la spectroscopie Raman.

Le plan est d'utiliser une analyse par composantes principales (PCA) avec le module SKlearn.

Context

To be written.

Database

Usage

You can do :

import ramandb

db = RamanDB()
matrix = db.getIntensities() 

to get th spectra. You should run the tests testDatabase.py to confirm everything is working. The database will be downloaded if needed.

Building the database

You do not need to do that. This is for reference.

It is easier to use a database to get all the spectra easily, with all their metadata. The first step is to create the database.

The database used is currently sqlite3, available on macOS by default, and downloadable for Windows. We open the (empty) database:

sqlite3 raman.db

We create the first table that will contain the name of the spectral files.

CREATE TABLE spectralfiles (path text, md5 text, fileId integer primary key autoincrement, date text);

Using a quick Unix command, we can get all files with their "hash" (a unique identifier). We format and then write to a csv file for import:

find . -name "*.txt" -exec md5 {} \; > files+md5.txt
perl formatmd5.pl < files+md5.txt > files.csv

Then, back into sqlite3 raman.db, we actually import:

.mode csv
.sep "|"
.import files.csv spectralfiles

Then we import all spectra. From the Terminal, we get the files and format them:

find . -name "*csv" -exec python3 formatspectroforimport.py {} \; > importall.csv

then import the files.

.mode csv
.sep "|"
.import files.csv importall.csv

Database schema

CREATE TABLE files (path text, md5 text primary key, date text );
CREATE TABLE spectra (wavelength real, intensity real, md5 text);
CREATE INDEX md5Idx on spectra(md5);
CREATE INDEX md5Idx2 on files(md5);
CREATE INDEX waveIdx on spectra(wavelength);
CREATE TABLE samples (name text, identifier text, type text, grape text, alcohol number, url text);