# O(N^2) DP approach
# runs in ~11s with pypy
# L'approche utilisee est assez simple et naturelle mais pourrait sembler naivement en O(N^3),
# la complexite en O(N^2) repose sur une prog. dyn. pour le calcul efficace des intersections.
# On ne detaille pas ici, on renvoie a l'explication claire et concise de (section 3.1) :
# Sand et al., Algorithms for Computing the Triplet and Quartet Distances for Binary and General Trees
# in Biology, 2013, http://www.cs.au.dk/~gerth/papers/biology13.pdf
## Input
S = raw_input().split()
N = len(S)
Sall = set(S)
A = {}
rosalib.parse_newick(raw_input(), A, True)
B = {}
rosalib.parse_newick(raw_input(), B, True)
## Fonctions de construction
def children(T, u, u0):
return (v for v in T[u] if v != u0)
# Pour chaque arete orientee e = (u0,u) de l'arbre T,
# S(e) = ensemble des feuilles du sous-arbre enracine en u
# pointe par la direction de e
# Calcul des S(e) (et numerotation des aretes au passage)
def dfs(T, u, u0, E, Enum, S):
if u[0] != '@': # leaf
def main():
T = {}
R = rosalib.parse_newick(input(),T,False)
P = dfs_prob(T,R)
# l'enonce les veut dans l'ordre AA,Aa,aa
print(' '.join(map(str,reversed(P))))
#!/usr/bin/env python3
import rosalib
# NB: "Given: An unrooted *binary* tree"
T = {}
R = rosalib.parse_newick(input(),T,False) # we do not unroot it
S = sorted(X for X in T if X[0]!='@') # named species (necessarily leaves?)
CharTbl = []
def dfs(u):
Su = set()
if u[0]!='@':
Su.add(u)
for v in T[u]:
Su |= dfs(v)
if 1<len(Su)<len(S)-1: # nontrivial char
CharTbl.append([X in Su for X in S])
return Su
dfs(R)
for L in CharTbl:
print(''.join('1' if x else '0' for x in L))
#!/usr/bin/env python3
import rosalib
T = {}
R = rosalib.parse_newick(input(), T, False)
F = rosalib.parse_fasta()
N = len(F[0][1]) # size of DNAs
DNA = {x: y for x, y in F}
def dfs(u, u0, i, si, ti):
if DNA[u][i] == si:
yield u
elif DNA[u][i] == ti:
for v in T[u]:
yield from dfs(v, u, i, si, ti)
for s in T:
for t in T[s]:
for i in range(N):
if DNA[s][i] != DNA[t][i]:
for w in dfs(t, s, i, DNA[s][i], DNA[t][i]):
print('%s %s %d %s->%s->%s' %
(t, w, i + 1, DNA[s][i], DNA[t][i], DNA[s][i]))
#!/usr/bin/env python3
import rosalib
A = ['A', 'C', 'G', 'T', '-']
tree = input()
T = {}
R = rosalib.parse_newick(tree, T, False)
F = rosalib.parse_fasta()
N = len(F[0][1]) # size of DNAs
DNA = {x: y for x, y in F}
# dp(i,u,a) = min sum of hamming distances of i-th letters on edges of
# the subtree rooted at node u using a in A as i-th letter
# of node u's DNA strand
# O(len(T) * N * len(A))
memo, pred = {}, {}
def dp(i, u, a):
assert (u not in DNA) # u is not a leaf
if (i, u, a) in memo:
return memo[i, u, a]
smin, pmin = 0, []
for v in T[u]:
if v in DNA: # v leaf
b = DNA[v][i]
sv, pv = int(a != b), b
else:
sv, pv = float('inf'), None
def main():
T = {}
R = rosalib.parse_newick(input(), T, False)
P = dfs_prob(T, R)
# l'enonce les veut dans l'ordre AA,Aa,aa
print(' '.join(map(str, reversed(P))))
#!/usr/bin/env python3
import rosalib
S = list(input().split())
S = {S[i]:i for i in range(len(S))}
Sall = frozenset(range(len(S)))
T1 = {}
R1 = rosalib.parse_newick(input(),T1,False) # we do not unroot it
T2 = {}
R2 = rosalib.parse_newick(input(),T2,False) # we do not unroot it
# similar to character table building from CTBL
def dfs(T,u,Splits):
if u[0]!='@':
return frozenset([S[u]])
Su = frozenset()
for v in T[u]:
Su |= dfs(T,v,Splits)
if 1<len(Su)<len(S)-1: # nontrivial char
if 0 in Su:
Splits.add(Su)
else:
Splits.add(Sall-Su) # on ajoute le complementaire
return Su
S1 = set()
dfs(T1,R1,S1)
S2 = set()
dfs(T2,R2,S2)
d = 2*(len(S)-3)-2*len(S1&S2)
#!/usr/bin/env python3
import rosalib
# NB: "Given: An unrooted *binary* tree"
T = {}
R = rosalib.parse_newick(input(), T, False) # we do not unroot it
S = sorted(X for X in T if X[0] != '@') # named species (necessarily leaves?)
CharTbl = []
def dfs(u):
Su = set()
if u[0] != '@':
Su.add(u)
for v in T[u]:
Su |= dfs(v)
if 1 < len(Su) < len(S) - 1: # nontrivial char
CharTbl.append([X in Su for X in S])
return Su
dfs(R)
for L in CharTbl:
print(''.join('1' if x else '0' for x in L))
#!/usr/bin/env python3
import rosalib
T = {}
R = rosalib.parse_newick(input(),T,False)
F = rosalib.parse_fasta()
N = len(F[0][1]) # size of DNAs
DNA = {x:y for x,y in F}
def dfs(u,u0,i,si,ti):
if DNA[u][i]==si:
yield u
elif DNA[u][i]==ti:
for v in T[u]:
yield from dfs(v,u,i,si,ti)
for s in T:
for t in T[s]:
for i in range(N):
if DNA[s][i]!=DNA[t][i]:
for w in dfs(t,s,i,DNA[s][i],DNA[t][i]):
print('%s %s %d %s->%s->%s' % (t,w,i+1,DNA[s][i],DNA[t][i],DNA[s][i]))
# O(N^2) DP approach
# runs in ~11s with pypy
# L'approche utilisee est assez simple et naturelle mais pourrait sembler naivement en O(N^3),
# la complexite en O(N^2) repose sur une prog. dyn. pour le calcul efficace des intersections.
# On ne detaille pas ici, on renvoie a l'explication claire et concise de (section 3.1) :
# Sand et al., Algorithms for Computing the Triplet and Quartet Distances for Binary and General Trees
# in Biology, 2013, http://www.cs.au.dk/~gerth/papers/biology13.pdf
## Input
S = raw_input().split()
N = len(S)
Sall = set(S)
A = {}
rosalib.parse_newick(raw_input(),A,True)
B = {}
rosalib.parse_newick(raw_input(),B,True)
## Fonctions de construction
def children(T,u,u0):
return (v for v in T[u] if v!=u0)
# Pour chaque arete orientee e = (u0,u) de l'arbre T,
# S(e) = ensemble des feuilles du sous-arbre enracine en u
# pointe par la direction de e
# Calcul des S(e) (et numerotation des aretes au passage)
def dfs(T,u,u0,E,Enum,S):
if u[0]!='@': # leaf
l = {u}