| cd my_huge_DCIM_folder | |
| for Y in {2017..2019} ; do | |
| echo "****** YEAR $Y ******" | |
| for M in {01..12} ; do | |
| echo "------ Month $M ------" | |
| mkdir -p "$Y/$M" | |
| mv IMG_${Y}${M}*.jpg "$Y/$M/" | |
| mv VID_${Y}${M}*.mp4 "$Y/$M/" | |
| mv PANO_${Y}${M}*.jpg "$Y/$M/" | |
| done |
| echo "Create a subset of ImageNet of 2 images per class for debug" | |
| SOURCE="data/ILSVRC2012/train" | |
| DEST="data/ILSVRC2012_samples/train" | |
| SOURCE="data/ILSVRC2012/validation" | |
| DEST="data/ILSVRC2012_samples/validation" | |
| cd "$SOURCE" | |
| for file in */ ; do |
Traducción ES: Framartin, 2019
Si nunca has jugado o si estás jugando con gente que nunca antes has jugado, coge dos cartas al azar y ponlas boca arriba en la mesa entre todos los jugadores.
Busque el símbolo idéntico entre las dos tarjetas (misma forma, mismo color, sólo el tamaño puede ser diferente). El primer jugador que encuentra el símbolo correcto lo nombra en voz alta y coge dos cartas nuevas que se colocan boca arriba sobre la mesa.
| # List of ISO 2 country codes of France and DOM-TOM (Overseas territories of France) | |
| # gathered from https://fr.wikipedia.org/wiki/France_d%27outre-mer#Liste | |
| COUNTRYCODES_FR_DOM_TOM = ['BL', 'FR', 'GF', 'GP', 'MF', 'MQ', 'NC', 'PF', 'PM', | |
| 'RE', 'TF', 'YT', 'WF'] |
| import sentry_sdk | |
| from scrapy.exceptions import NotConfigured | |
| class SentryLogging(object): | |
| """ | |
| Send exceptions and errors to Sentry. | |
| """ | |
| @classmethod | |
| def from_crawler(cls, crawler): |
| def node_generator(node): | |
| """ | |
| Generator that takes an Esprima object (or a Esprima node) from the esprima | |
| module converted as a dict, and outputs all child nodes at any level of the | |
| tree. It's useful to browse the entire tree. | |
| Subnodes are generated by browsing all the keys. It's not the most | |
| optimized way to browse the tree, because some keys will never contains | |
| child nodes. But it's very simple, and you're sure to not miss any subnodes. | |
| """ | |
| if node: # not empty dict or list |
| library(spdep) | |
| data(oldcol) | |
| lw <- nb2listw(COL.nb) | |
| attr(lw, "region.id") | |
| region.id.wanted.order <- attr(lw, "region.id")[c(1:41, 44:49, 42, 43)] # cannot be obtain with subset.listw | |
| # we convert it to a Matrix and re-normalized it with mat2listw() | |
| W <- as(lw, "CsparseMatrix") | |
| W <- W[region.id.wanted.order, region.id.wanted.order] | |
| style <- lw$style | |
| lw1 <- mat2listw(W, row.names = region.id.wanted.order, style = style) # re-normalize to keep the style |
| K <- ifelse(colnames(X)[1] == "(Intercept)", 2, 1) | |
| m <- ncol(X) | |
| # check if there are enough regressors | |
| if (m > 1) { | |
| WX <- matrix(nrow=nrow(X),ncol=(m-(K-1))) | |
| for (k in K:m) { | |
| wx <- lag.listw(listw, X[,k], | |
| zero.policy=zero.policy) | |
| if (any(is.na(wx))) | |
| stop("NAs in lagged independent variable") |