X-Git-Url: https://git.auder.net/?a=blobdiff_plain;f=TODO;h=3c1fd78ec8aadcc1ff7a9df076932c2db515f99b;hb=dc1aa85a96bbf815b0d896c22a9b4a539a9e8a9c;hp=662635d48ac9904c026063d48781fb4f8c4d01dc;hpb=7709d507dfab9256a401f2c77ced7bc70d90fec3;p=epclust.git

diff --git a/TODO b/TODO
index 662635d..3c1fd78 100644
--- a/TODO
+++ b/TODO
@@ -15,9 +15,54 @@ geometric structure of high dim data and dim reduction 2011
 https://docs.docker.com/engine/getstarted/step_one/
 
 A faire:
- - finir les experiences (sur nb de classes, nb de curves / chunk, nb de procs) 
+ - finir les experiences (sur nb de classes, nb de curves / chunk, nb de procs)
    et sur d'autres architectures
- - interface matrice -> binaire
+
+dans old_C_code/build :
+cmake ../stage1/src
+make
+
+dans data/, lancer R puis :
+source("../old_C_code/wrapper.R")
+serialize("../old_C_code/build", "2009.csv","2009.bin",1)
+library(parallel)
+np = detectCores()
+nbSeriesPerChunk = 3000
+nbClusters = 20
+ppam_exe("../old_C_code/build",np,"2009.bin",nbSeriesPerChunk,nbClusters)
+C = getMedoids("../old_C_code/build", "ppamResult.xml", "ppamFinalSeries.bin")
+first100series = deserialize("../old_C_code/build", "2009.bin", "2009.csv.part", "1-100")
+distor = getDistor("../old_C_code/build", "ppamResult.xml", "2009.bin")
+
+- interface matrice -> binaire
+	 OK
+
  - courbe synchrone
+	 ??
+
+Piste à explorer pour les comparaisons: H20
+
+renvoyer nombre d'individues par classe ? (+ somme ?)
+hypothèse : données déjà ordonnées 48 1/2H sur 365j
+utiliser du mixmod avec modèles allongés
+doit toutner sur machine plutôt standard, utilisateur "lambda"
+utiliser Rcpp ?
+
+=====
+
+trategies for upscaling
+From 25K to 25M : in 1000 chunks of 25K
+Reference values :
+ K 0 = 200 super consumers (SC)
+ K ∗ = 15 nal clusters
+1st strategy
+ Do 1000 times ONLY Energycon's 1st-step strategy on 25K clients
+ With the 1000 × K 0 SC perform a 2-step run leading to K ∗ clusters
+
+--> il faut s'arranger pour que 
 
-Piste à explorer pour les comparaisons: H20 
+2nd strategy
+ Do 1000 times Energycon's 2-step strategy on 25K clients leading to
+ 1000 × K ∗ intermediate clusters
+ Treat the intermediate clusters as individual curves and perform a
+ single 2-step run to get K ∗ nal clusters