with window analysis

empath/core.py

@@ -27,6 +27,10 @@ def load(self,file):
                     self.cats[name].append(t)
                     #self.invcats[t].append(name)
 
+    def analyze_term_window(self,doc,targets,categories=None,window_size=10,normalize=False):
+        tokenizer = util.window_tokenizer(window_size,targets)
+        return self.analyze(doc,categories,tokenizer,normalize)
+
     def analyze(self,doc,categories=None,tokenizer="default",normalize=False):
         if isinstance(doc,list):
             doc = "\n".join(doc)

empath/data/categories.tsv

@@ -190,5 +190,5 @@ giving	bring	generous	thank	favor	save	return	generously	redeem	earn	gift	thankf
 contentment	languid	pleaser	closeness	pulsing	fullness	happiness	gentleness	nostalgia	fervor	pleasing	pleasurable	amorous	compassion	tantalizing	kindness	thrill	profound	delightful	delightfully	sensation	arouse	momentary	familiarity	exhilaration	passion	gratitude	joy	relaxation	lightness	bask	fondness	deliciously	certainty	sweetness	feel	satisfaction	remembrance	moan	contentment	bliss	tranquillity	tranquility	appreciation	orgasm	warmth	excitement	feeling	affection	coolness	contented	fulfillment	genuine	savor	pure	enjoyment	delight	pleasure	calmness	relief	intimacy	gratification
 writing	essay	poetry	paper	thesis	portfolio	creative	parchment	write	calligraphy	tablet	writing	blackboard	clipboard	stationery	brainstorm	homework	descriptive	notation	read	eulogy	diary	postcard	list	summary	revision	page	wrote	lettering	excerpt	scribe	compile	font	publish	pen	paragraph	assignment	print	reading	email	blurb	handwritten	jot	booklet	journal	binder	article	scribbling	planner	informational	revise	manuscript	pamphlet	marker	poem	sheet	script	transcript	handbook	typing	pencil	paperwork	paperback	detail	brochure	copy	word	instruction	chalk	printing	neat	editorial	newspaper	handwriting	prompt	rewrite	file	ink	autobiography	typed	author	alphabet	note	book	composition	typewriter	quote	notebook	printed	letter	bulletin	entry
 rural	ramshackle	barren	cornfield	schoolhouse	quaint	plantation	townsfolk	folk	forested	idyllic	scenic	farmhouse	meadow	uninhabited	prairie	migration	southwest	settlement	farm	wheat	outlying	farmland	environment	settler	road	secluded	colony	county	village	campground	landscape	vineyard	west	locally	lakeside	live	acre	farming	farmer	wildlife	landscaping	western	cottage	valley	heartland	bungalow	property	tractor	grassy	pasture	cultivate	picturesque	woodland	isolated	barn	province	agricultural	countryside	resident	grove	ranch	south	grassland	rural	inhabitant
-positive_emotion happiness	enlighten	better	enthusiasm	pride	joyful	compassion	dearly	forgiving	kindness	bravery	closure	thrill	honestly	triumph	bond	honesty	alive	concern	reunite	joy	surprise	forgiveness	assurance	sympathize	understanding	reason	rejoice	care	faith	great	empathy	certainty	keep	trustworthy	affection	cherish	emotion	love	family	trusting	respect	trust	gratitude	confidence	adoration	friend	happy	overjoyed	determination	reassurance	glad	loved	admiration	wish	accomplishment	optimism	excitement	convince	hope	freedom	feeling	eagerness	willingness	sincere	sincerity	honest	genuine	comfort	elation	thrilled	loyalty	curiosity	unconditionally	proud
+positive_emotion	happiness	enlighten	better	enthusiasm	pride	joyful	compassion	dearly	forgiving	kindness	bravery	closure	thrill	honestly	triumph	bond	honesty	alive	concern	reunite	joy	surprise	forgiveness	assurance	sympathize	understanding	reason	rejoice	care	faith	great	empathy	certainty	keep	trustworthy	affection	cherish	emotion	love	family	trusting	respect	trust	gratitude	confidence	adoration	friend	happy	overjoyed	determination	reassurance	glad	loved	admiration	wish	accomplishment	optimism	excitement	convince	hope	freedom	feeling	eagerness	willingness	sincere	sincerity	honest	genuine	comfort	elation	thrilled	loyalty	curiosity	unconditionally	proud
 musical	chord	producer	radio	keyboard	clarinet	ballet	harmonica	glee	instrument	production	rhythm	piano	songwriting	string	song	songwriter	band	banjo	artist	audience	saxophone	acoustic	gospel	quartet	auditorium	concert	turntable	medley	melody	harp	album	chorus	rehearse	orchestra	lyric	violinist	rendition	solo	bass	compilation	guitar	tuner	genre	sing	impromptu	lyrical	soundtrack	drumming	drum	talented	musically	cd	composer	singing	classical	theme	interlude	gig	hymn	symphony	music	upbeat	ballad	melodic	hum	musician	singer	instrumental	burlesque	soprano	vocal	video	duet	performer	classic	soloist	sax	tenor	violin	trumpet	mainstream	dancing	perform	guitarist	actor	note	freestyle	drummer	harmony	performance	pianist	jamming	composition	reggae	play	jazz	theatre	collaboration	flute	rehearsal	audition	vocalist	tune	nirvana	recital	playing	riff	bassist	audio	musical

empath/helpers.py

@@ -1,3 +1,18 @@
+from itertools import islice
+
+def partitions(l, n):
+    for i in range(0, len(l), n):
+        yield l[i:i+n]
+
+def window(seq, n):
+    it = iter(seq)
+    result = tuple(islice(it, n))
+    if len(result) == n:
+        yield result    
+    for elem in it:
+        result = result[1:] + (elem,)
+        yield result
+
 def default_tokenizer(doc):
     return doc.split()
 
@@ -7,3 +22,13 @@ def bigram_tokenizer(doc):
         yield t1
         yield t1+"_"+t2
     yield tokens[-1]
+
+def window_tokenizer(window_size, targets):
+    def window_func(doc,target_dic={t:t for t in targets}):
+        tokens = doc.split()
+        print(tokens)
+        for w in partitions(tokens, window_size):
+            if any([x in target_dic for x in w]):
+                for w_ in w: 
+                    yield w_
+    return window_func