search

davidhallac / NetworkLasso

22 stars 19 forks source link

cvxpy.error.SolverError: Solver 'ECOS' failed #1

Closed hsong60csmile closed 6 years ago

hsong60csmile commented 6 years ago

When I run housing_paper.py in Linux, I got the following error:

"Solver '%s' failed. Try another solver." % solver.name()) cvxpy.error.SolverError: Solver 'ECOS' failed. Try another solver.

Can anyone share the solutions? Thanks very much!

davidhallac commented 6 years ago

Hmm, this appears to be a difference in the way that CVXPY handles poorly-scaled problems. While it used to return "None" for such problems, now it appears to throw an error: https://github.com/cvxgrp/cvxpy/issues/103.

The best workaround would be to wrap line 44 in the code with a "try... catch..." statement, like this:

try: result = p.solve() except SolverError: result = p.solve(solver=SCS)

hsong60csmile commented 6 years ago

Thanks for your reply. However, with your suggestion, it seems it doesn't work.

It happens when the lambda is 0.1 and it's normal while lambda is 0. I'm not sure if that's because the problem badly scaled as: https://groups.google.com/forum/#!topic/cvxpy/x7KdlSF6Ts0

As the test in windows failed, I run the code on Bush on windows along with all tests successfully. But 'nondistributed_svm_comparison.py' doesn't have this problem. Very strange.

On Thu, Sep 21, 2017 at 10:22 PM, davidhallac notifications@github.com wrote:

Hmm, this appears to be a difference in the way that CVXPY handles poorly-scaled problems. While it used to return "None" for such problems, now it appears to throw an error: cvxgrp/cvxpy#103 https://github.com/cvxgrp/cvxpy/issues/103.

The best workaround would be to wrap line 44 in the code with a "try... catch..." statement, like this:

try: result = p.solve() except SolverError: result = p.solve(solver=SCS)

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/davidhallac/NetworkLasso/issues/1#issuecomment-331140168, or mute the thread https://github.com/notifications/unsubscribe-auth/AbBQjq4E_N-C59kmEnQ8cSc10I1ky_sZks5sklUHgaJpZM4Pes43 .

davidhallac commented 6 years ago

Yes, I think that the issue is with the scaling of the problem. However, the workaround I mentioned before should fix that (since we are switching from ECOS to SCS as the underlying solver). Do you still get an error when you run the "try... except..." code? What error are you getting?

hsong60csmile commented 6 years ago

Thanks for your reply.

It said "except SolverError: SyntaxError: invalid syntax". As I'm not familiar with Python, probably I will need time to fix it.

On Thu, Sep 28, 2017 at 12:39 AM, davidhallac notifications@github.com wrote:

Yes, I think that the issue is with the scaling of the problem. However, the workaround I mentioned before should fix that (since we are switching from ECOS to SCS as the underlying solver). Do you still get an error when you run the "try... except..." code? What error are you getting?

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/davidhallac/NetworkLasso/issues/1#issuecomment-332543321, or mute the thread https://github.com/notifications/unsubscribe-auth/AbBQjvTKN2rVK_fl7qpsdSrA3DADTkLZks5sml4mgaJpZM4Pes43 .

davidhallac commented 6 years ago

That appears to be a Syntax error, so it probably has to do with the code (and not an ECOS error). Can you post the chunk of code that you used?

hsong60csmile commented 6 years ago

Thanks for your reply.

I have solved that problem. However, it still has the same problem:

Traceback (most recent call last): File "housing_paper.py", line 456, in main() File "housing_paper.py", line 375, in main result = p.solve(verbose=False) File "/home/songhui/anaconda2/envs/stanford/lib/python2.7/site-packages/cvxpy/problems/problem.py", line 209, in solve return self._solve(*args, **kwargs) File "/home/songhui/anaconda2/envs/stanford/lib/python2.7/site-packages/cvxpy/problems/problem.py", line 335, in _solve self._update_problem_state(results_dict, sym_data, solver) File "/home/songhui/anaconda2/envs/stanford/lib/python2.7/site-packages/cvxpy/problems/problem.py", line 447, in _update_problem_state "Solver '%s' failed. Try another solver." % solver.name()) cvxpy.error.SolverError: Solver 'ECOS' failed. Try another solver.

It seems that the changed part didn't execute. I attached the code as follows. The only difference from the original is from line 44 to 47.

Thanks very much!

On Thu, Sep 28, 2017 at 2:43 PM, davidhallac notifications@github.com wrote:

That appears to be a Syntax error, so it probably has to do with the code (and not an ECOS error). Can you post the chunk of code that you used?

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/davidhallac/NetworkLasso/issues/1#issuecomment-332726281, or mute the thread https://github.com/notifications/unsubscribe-auth/AbBQjqXDYqZxlN4QMNmt3vIeafX_4GT6ks5smyP7gaJpZM4Pes43 .

from snap import from cvxpy import import numpy as np from numpy import linalg as LA import math from multiprocessing import Pool

Plotting

import csv import os
import tempfile os.environ['MPLCONFIGDIR'] = tempfile.mkdtemp() import matplotlib matplotlib.use('Agg') matplotlib.rc('text',usetex=True) import matplotlib.pyplot as plt

Other function in this folder

from z_u_solvers import solveZ, solveU import sys

def solveX(data): inputs = int(data[data.size-1]) lamb = data[data.size-2] rho = data[data.size-3] sizeData = data[data.size-4] mu = data[data.size-5] x = data[0:inputs] a = data[inputs:(inputs + sizeData)] neighs = data[(inputs + sizeData):data.size-5] xnew = Variable(inputs,1)

#Fill in objective function here! Params: Xnew (unknown), a (side data at node)
g = square(xnew[0]*a[0] + xnew[1]*a[1] + xnew[2]*a[2] + xnew[3] - a[4]) + mu*(square(xnew[0]) + square(xnew[1]) + square(xnew[2]))

h = 0
for i in range(neighs.size/(2*inputs+1)):
    weight = neighs[i*(2*inputs+1)]
    if(weight != 0):
        u = neighs[i*(2*inputs+1)+1:i*(2*inputs+1)+(inputs+1)]
        z = neighs[i*(2*inputs+1)+(inputs+1):(i+1)*(2*inputs+1)]
        h = h + rho/2*square(norm(xnew - z + u))
objective = Minimize(50*g+50*h)
constraints = []
p = Problem(objective, constraints)
try:
    result = p.solve()
except SolverError:
    result = p.solve(solver=SCS)
#result = p.solve()
if(result == None):
    #CVXOPT scaling issue. Rarely happens (but occasionally does when running thousands of tests)
    objective = Minimize(51*g+52*h)
    p = Problem(objective, constraints)
    result = p.solve(verbose=False)
    if(result == None):
        print "SCALING BUG"
        objective = Minimize(52*g+50*h)
        p = Problem(objective, constraints)
        result = p.solve(verbose=False)
return xnew.value, g.value

def runADMM(G1, sizeOptVar, sizeData, lamb, rho, numiters, x, u, z, a, edgeWeights, useConvex, epsilon, mu):

nodes = G1.GetNodes()
edges = G1.GetEdges()

maxNonConvexIters = 6*numiters

#Find max degree of graph; hash the nodes
(maxdeg, counter) = (0, 0)
node2mat = TIntIntH()
for NI in G1.Nodes():
    maxdeg = max(maxdeg, NI.GetDeg())
    node2mat.AddDat(NI.GetId(), counter)
    counter = counter + 1

#Stopping criteria
eabs = math.pow(10,-2)
erel = math.pow(10,-3)
(r, s, epri, edual, counter) = (1,1,0,0,0)
A = np.zeros((2*edges, nodes))
for EI in G1.Edges():
    A[2*counter,node2mat.GetDat(EI.GetSrcNId())] = 1
    A[2*counter+1, node2mat.GetDat(EI.GetDstNId())] = 1
    counter = counter+1
(sqn, sqp) = (math.sqrt(nodes*sizeOptVar), math.sqrt(2*sizeOptVar*edges))

#Non-convex case - keeping track of best point so far
bestx = x
bestu = u
bestz = z
bestObj = 0
cvxObj = 10000000*np.ones((1, nodes))
if(useConvex != 1):
    #Calculate objective
    for i in range(G1.GetNodes()):
        bestObj = bestObj + cvxObj[0,i] 
    for EI in G1.Edges():
        weight = edgeWeights.GetDat(TIntPr(EI.GetSrcNId(), EI.GetDstNId()))
        edgeDiff = LA.norm(x[:,node2mat.GetDat(EI.GetSrcNId())] - x[:,node2mat.GetDat(EI.GetDstNId())])
        bestObj = bestObj + lamb*weight*math.log(1 + edgeDiff / epsilon)
    initObj = bestObj

#Run ADMM
iters = 0
maxProcesses =  80
pool = Pool(processes = min(max(nodes, edges), maxProcesses))
while(iters < numiters and (r > epri or s > edual or iters < 1)):

    #x-update
    neighs = np.zeros(((2*sizeOptVar+1)*maxdeg,nodes))
    edgenum = 0
    numSoFar = TIntIntH()
    for EI in G1.Edges():
        if (not numSoFar.IsKey(EI.GetSrcNId())):
            numSoFar.AddDat(EI.GetSrcNId(), 0)
        counter = node2mat.GetDat(EI.GetSrcNId())
        counter2 = numSoFar.GetDat(EI.GetSrcNId())
        neighs[counter2*(2*sizeOptVar+1),counter] = edgeWeights.GetDat(TIntPr(EI.GetSrcNId(), EI.GetDstNId()))
        neighs[counter2*(2*sizeOptVar+1)+1:counter2*(2*sizeOptVar+1)+(sizeOptVar+1),counter] = u[:,2*edgenum] 
        neighs[counter2*(2*sizeOptVar+1)+(sizeOptVar+1):(counter2+1)*(2*sizeOptVar+1),counter] = z[:,2*edgenum]
        numSoFar.AddDat(EI.GetSrcNId(), counter2+1)

        if (not numSoFar.IsKey(EI.GetDstNId())):
            numSoFar.AddDat(EI.GetDstNId(), 0)
        counter = node2mat.GetDat(EI.GetDstNId())
        counter2 = numSoFar.GetDat(EI.GetDstNId())
        neighs[counter2*(2*sizeOptVar+1),counter] = edgeWeights.GetDat(TIntPr(EI.GetSrcNId(), EI.GetDstNId()))
        neighs[counter2*(2*sizeOptVar+1)+1:counter2*(2*sizeOptVar+1)+(sizeOptVar+1),counter] = u[:,2*edgenum+1] 
        neighs[counter2*(2*sizeOptVar+1)+(sizeOptVar+1):(counter2+1)*(2*sizeOptVar+1),counter] = z[:,2*edgenum+1]
        numSoFar.AddDat(EI.GetDstNId(), counter2+1)

        edgenum = edgenum+1

    temp = np.concatenate((x,a,neighs,np.tile([mu, sizeData,rho,lamb,sizeOptVar], (nodes,1)).transpose()), axis=0)
    values = pool.map(solveX, temp.transpose())
    newx = np.array(values)[:,0].tolist()
    newcvxObj = np.array(values)[:,1].tolist()
    x = np.array(newx).transpose()[0]
    cvxObj = np.reshape(np.array(newcvxObj), (-1, nodes))

    #z-update
    ztemp = z.reshape(2*sizeOptVar, edges, order='F')
    utemp = u.reshape(2*sizeOptVar, edges, order='F')
    xtemp = np.zeros((sizeOptVar, 2*edges))
    counter = 0
    weightsList = np.zeros((1, edges))
    for EI in G1.Edges():
        xtemp[:,2*counter] = np.array(x[:,node2mat.GetDat(EI.GetSrcNId())])
        xtemp[:,2*counter+1] = x[:,node2mat.GetDat(EI.GetDstNId())]
        weightsList[0,counter] = edgeWeights.GetDat(TIntPr(EI.GetSrcNId(), EI.GetDstNId()))
        counter = counter+1
    xtemp = xtemp.reshape(2*sizeOptVar, edges, order='F')
    temp = np.concatenate((xtemp,utemp,ztemp,np.reshape(weightsList, (-1,edges)),np.tile([epsilon, useConvex, rho,lamb,sizeOptVar], (edges,1)).transpose()), axis=0)
    newz = pool.map(solveZ, temp.transpose())
    ztemp = np.array(newz).transpose()[0]
    ztemp = ztemp.reshape(sizeOptVar, 2*edges, order='F')
    s = LA.norm(rho*np.dot(A.transpose(),(ztemp - z).transpose())) #For dual residual
    z = ztemp

    #u-update
    (xtemp, counter) = (np.zeros((sizeOptVar, 2*edges)), 0)
    for EI in G1.Edges():
        xtemp[:,2*counter] = np.array(x[:,node2mat.GetDat(EI.GetSrcNId())])
        xtemp[:,2*counter+1] = x[:,node2mat.GetDat(EI.GetDstNId())]
        counter = counter + 1
    temp = np.concatenate((u, xtemp, z, np.tile(rho, (1,2*edges))), axis=0)
    newu = pool.map(solveU, temp.transpose())
    u = np.array(newu).transpose()

    #Update best objective (for non-convex)
    if(useConvex != 1):
        tempObj = 0
        #Calculate objective
        for i in range(G1.GetNodes()):
            tempObj = tempObj + cvxObj[0,i]
        initTemp = tempObj
        for EI in G1.Edges():
            weight = edgeWeights.GetDat(TIntPr(EI.GetSrcNId(), EI.GetDstNId()))
            edgeDiff = LA.norm(x[:,node2mat.GetDat(EI.GetSrcNId())] - x[:,node2mat.GetDat(EI.GetDstNId())])
            tempObj = tempObj + lamb*weight*math.log(1 + edgeDiff / epsilon)
        #Update best variables
        if(tempObj <= bestObj):
            bestx = x
            bestu = u
            bestz = z
            bestObj = tempObj
            print "Iteration ", iters, "; Obj = ", tempObj, "; Initial = ", initTemp

        if(iters == numiters - 1 and numiters < maxNonConvexIters):
            if(bestObj == initObj):
                numiters = numiters+1

    #Stopping criterion - p19 of ADMM paper
    epri = sqp*eabs + erel*max(LA.norm(np.dot(A,x.transpose()), 'fro'), LA.norm(z, 'fro'))
    edual = sqn*eabs + erel*LA.norm(np.dot(A.transpose(),u.transpose()), 'fro')
    r = LA.norm(np.dot(A,x.transpose()) - z.transpose(),'fro')
    s = s

    #print r, epri, s, edual
    iters = iters + 1

pool.close()
pool.join()

if(useConvex == 1):
    return (x,u,z,0,0)
else:
    return (bestx,bestu,bestz,0,0)

def main():

#Set parameters
useConvex = 1
rho = 0.001
numiters = 50
thresh = 1000#10000
lamb = 0.0
startVal = 0.1#0.01
useMult = 1 #1 for mult, 0 for add
addUpdateVal = 0.1 
multUpdateVal = 1.25#1.1

mu = 0.5 #For LS regularization
#Test/Validation Set Information
numNeighs = 5 #For data we keep
testSetSize = 200
validationSetSize = 0
numNewNeighs = 5 #For test/validation nodes
#Size of x
sizeOptVar = 4
#Size of side information at each node
sizeData = 5
#Non-convex vars
epsilon = 0.01

#Generate graph, edge weights
file = open("Data/Sacramentorealestatetransactions_Normalized.csv", "rU")
file.readline() #ignore first line
G1 = TUNGraph.New()
locations = TIntFltPrH()
dataset = TIntFltVH()
counter = 0
for line in file:
    G1.AddNode(counter)
    temp = TFltPr(float(line.split(",")[10]),float(line.split(",")[11]))
    locations.AddDat(counter, temp)
    tempData = TFltV()
    tempData.Add(float(line.split(",")[4]))
    tempData.Add(float(line.split(",")[5]))
    tempData.Add(float(line.split(",")[6]))
    if(line.split(",")[7] == "Residential"):
        tempData.Add(1)
    elif(line.split(",")[7] == "Condo"):
        tempData.Add(2)
    elif(line.split(",")[7] == "Multi-Family"):
        tempData.Add(3)
    else:
        tempData.Add(4)
    tempData.Add(float(line.split(",")[12])*10) #12 for normalized; 9 for raw
    dataset.AddDat(counter, tempData)
    counter = counter + 1

#Remove random subset of nodes for test and validation sets
testList = TIntV()
for i in range(testSetSize):
    temp = G1.GetRndNId()
    G1.DelNode(temp)
    testList.Add(temp)

validationList = TIntV()
for i in range(validationSetSize):
    temp = G1.GetRndNId()
    G1.DelNode(temp)
    validationList.Add(temp)

#For each node, find closest neightbors and add edge, weight = 5/distance
edgeWeights = TIntPrFltH()
for NI in G1.Nodes():
    distances = TIntFltH()
    lat1 = locations.GetDat(NI.GetId()).GetVal1()
    lon1 = locations.GetDat(NI.GetId()).GetVal2()
    for NI2 in G1.Nodes():
        if(NI.GetId() != NI2.GetId()):
            lat2 = locations.GetDat(NI2.GetId()).GetVal1()
            lon2 = locations.GetDat(NI2.GetId()).GetVal2()
            dlon = math.radians(lon2 - lon1)
            dlat = math.radians(lat2 - lat1)
            a2 = math.pow(math.sin(dlat/2),2) + math.cos(lat1)*math.cos(lat2) * math.pow(math.sin(dlon/2),2)
            c = 2 * math.atan2( math.sqrt(a2), math.sqrt(1-a2) ) 
            dist = 3961 * c
            distances.AddDat(NI2.GetId(), dist)

    distances.Sort(False, True)
    it = distances.BegI()
    for j in range(numNeighs):
        if (not G1.IsEdge(NI.GetId(), it.GetKey())):
            G1.AddEdge(NI.GetId(), it.GetKey())
            #Add edge weight
            temp = TIntPr(min(NI.GetId(), it.GetKey()), max(NI.GetId(), it.GetKey()))
            edgeWeights.AddDat(temp, 1/(it.GetDat()+ 0.1))
        it.Next()       

nodes = G1.GetNodes()
edges = G1.GetEdges()
print nodes, edges
print GetBfsFullDiam(G1, 1000, False);

# SaveEdgeList(G1, 'mygraph.txt')
# sys.exit()

#Get side information
a = np.zeros((sizeData, nodes))
counter = 0
for NI in G1.Nodes():
    a[0,counter] = dataset.GetDat(NI.GetId())[0]
    a[1,counter] = dataset.GetDat(NI.GetId())[1]
    a[2,counter] = dataset.GetDat(NI.GetId())[2]
    a[3,counter] = dataset.GetDat(NI.GetId())[3]
    a[4,counter] = dataset.GetDat(NI.GetId())[4]
    counter = counter + 1

#Initialize variables to 0
x = np.zeros((sizeOptVar,nodes))
u = np.zeros((sizeOptVar,2*G1.GetEdges()))
z = np.zeros((sizeOptVar,2*G1.GetEdges()))

# avgPrice = np.mean(a[4,:])
# print avgPrice

#Run regularization path
[plot1, plot2, plot3] = [TFltV(), TFltV(), TFltV()]
while(lamb <= thresh or lamb == 0):
    (x, u, z, pl1, pl2) = runADMM(G1, sizeOptVar, sizeData, lamb, rho + math.sqrt(lamb), numiters, x, u ,z, a, edgeWeights, useConvex, epsilon, mu)
    print "Lambda =", lamb
    mse = 0
    #Calculate accuracy on test set
    for i in testList:
        #Find closest neighbors
        distances = TIntFltH()
        lat1 = locations.GetDat(i).GetVal1()
        lon1 = locations.GetDat(i).GetVal2()
        counter = 0
        for NI2 in G1.Nodes():
            lat2 = locations.GetDat(NI2.GetId()).GetVal1()
            lon2 = locations.GetDat(NI2.GetId()).GetVal2()
            dlon = math.radians(lon2 - lon1)
            dlat = math.radians(lat2 - lat1)
            a2 = math.pow(math.sin(dlat/2),2) + math.cos(lat1)*math.cos(lat2) * math.pow(math.sin(dlon/2),2)
            c = 2 * math.atan2( math.sqrt(a2), math.sqrt(1-a2) ) 
            dist = 3961 * c
            distances.AddDat(counter, dist)     
            counter = counter + 1
        distances.Sort(False, True)

        #Predict price - CVXPY method
        xpred = Variable(sizeOptVar,1)
        g = 0
        it = distances.BegI()
        for j in range(numNewNeighs):
            weight = 1/(it.GetDat()+ 0.1)
            g = g + weight*norm(xpred - x[:, it.GetKey()])
            it.Next()   
        objective = Minimize(g)
        constraints = []
        p = Problem(objective, constraints)
        result = p.solve(verbose=False) 
        xpred = xpred.value

        #Find MSE
        regressors = dataset.GetDat(i)
        prediction = xpred[0]*float(regressors[0]) + xpred[1]*float(regressors[1]) + xpred[2]*float(regressors[2]) + xpred[3]
        mse = mse + math.pow(prediction - float(dataset.GetDat(i)[4]), 2)/testSetSize

    cons = 0
    for i in range(edges):
        if(np.all(z[:,2*i] == z[:,2*i + 1])):
            cons = cons + 1
    consensus = cons / float(edges)

    print mse, "= mse", consensus, " = consensus"
    plot1.Add(lamb)
    plot2.Add(mse)
    plot3.Add(consensus)
    if(lamb == 0):
        lamb = startVal
    elif(useMult == 1):
        lamb = lamb*multUpdateVal
    else:
        lamb = lamb + addUpdateVal

#Print/Save plot of results
if(thresh > 0):
    pl1 = np.array(plot1)
    pl2 = np.array(plot2)
    plt.plot(pl1, pl2/100) #To get to unit variance
    plt.xscale('log')
    plt.xlabel(r'$\lambda$')
    plt.ylabel('MSE')
    if(useConvex == 1):
        plt.savefig('image_housing_convex',bbox_inches='tight')
    else:
        plt.savefig('image_housing_nonconvex',bbox_inches='tight')

    #Plot of clustering
    plt.figure()
    pl3 = np.array(plot3)
    plt.plot(pl1, pl3)
    plt.xscale('log')
    plt.xlabel(r'$\lambda$')
    plt.ylabel('\% of edges in Consensus')

    if(useConvex == 1):
        plt.savefig('consensus_housing_convex',bbox_inches='tight')
        np.savetxt('text_housing_convex.out', (pl1, pl2, pl3), delimiter=',', fmt='%1.4f')
    else:
        plt.savefig('consensus_housing_nonconvex',bbox_inches='tight')
        np.savetxt('text_housing_nonconvex.out', (pl1, pl2, pl3), delimiter=',', fmt='%1.4f')

#Get data for heatmap
f = open('heatmap.txt','w')
xplot = np.zeros((sizeOptVar,nodes))
for i in range(nodes):
    for j in range(sizeOptVar):
        maxval = max(x[j,:])
        minval = min(x[j,:])
        xplot[j,i] = (x[j,i] - minval)/(maxval-minval + 0.01) * 512 - 128

        #numsds = 2
        #(avgval, stddev) = (average(x[j,:]), std(x[j,:])
        #xplot[j,i] = (((x[j,i] - avgval)/stddev)+numsds/2) / numsds

    newval = "{0:02x}{1:02x}{2:02x}".format(int(max(0, min(xplot[3,i], 255))), int(max(0, min(xplot[2,i], 255))), int(max(0, min(xplot[0,i], 255))))
    print "\"", newval, "\","
    f.write('\"' + newval + '\",')
f.close()

f = open('latlong.txt', 'w')
for NI in G1.Nodes():
    lat = locations.GetDat(NI.GetId()).GetVal1()
    lon = locations.GetDat(NI.GetId()).GetVal2()
    f.write('new GLatLng(' + str(lat) + ',' + str(lon) + '),' )
f.close()

if name == 'main': main()

davidhallac commented 6 years ago

Hmm, it seems to be working on my machine... Can you try replacing all the "p.solve()" lines with "p.solve(solver=SCS)"? So that means going back to the original code and changing lines 44, 49, and 54

hsong60csmile commented 6 years ago

Thanks for your reply.

After I change it, it has the solver issue from lambda=0. My friend in my lab tried this, it also has the solver problem.

On Fri, Sep 29, 2017 at 2:06 AM, davidhallac notifications@github.com wrote:

Hmm, it seems to be working on my machine... Can you try replacing all the "p.solve()" lines with "p.solve(solver=SCS)"? So that means going back to the original code and changing lines 44, 49, and 54

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/davidhallac/NetworkLasso/issues/1#issuecomment-332884830, or mute the thread https://github.com/notifications/unsubscribe-auth/AbBQjhUUVJznh8gxkoIltYZxLM4uUw49ks5sm8QHgaJpZM4Pes43 .

davidhallac commented 6 years ago

It gives an ECOS error when you use SCS as the solver? That can't be the case, since the code doesn't even call ECOS anymore...

Can you repost your updated code, along with the exact error you are getting when you run it?

hsong60csmile commented 6 years ago

Thanks for your reply.

It has been solved.

Here are the changes: line 44:result = p.solve(Solver=SCS) line 49: result = p.solve(Solver=SCS,verbose=False) line 54: result = p.solve(Solver=SCS,verbose=False) line 370: result = p.solve(Solver=SCS,verbose=False)

Previously, it didn't work because I didn't find line 370.

I really appreciate your help.

On Sat, Sep 30, 2017 at 8:15 AM, davidhallac notifications@github.com wrote:

It gives an ECOS error when you use SCS as the solver? That can't be the case, since the code doesn't even call ECOS anymore...

Can you repost your updated code, along with the exact error you are getting when you run it?

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/davidhallac/NetworkLasso/issues/1#issuecomment-333251697, or mute the thread https://github.com/notifications/unsubscribe-auth/AbBQjkvhe5JENN07ycr5zLFudSOn5A9Kks5snWv6gaJpZM4Pes43 .