Faulty output in task01

[wasanti]

Hello, I am using the FESTIM workshop on debian 11 with FESTIM 1.3.1 installed via pip and fenics via apt version : fenics/oldstable,now 1:2019.2.0.5 amd64 [installed].

The example (task01) does not raise an error, however, the "mobile_concentration.txt" output file contains one line repeated for every step. The value evolves as expected with each timestep.

Is this a known bug with my specific version? Any suggestions on how to further debug are wellcome.

[RemDelaporteMathurin]

Hi! Is this a case of https://festim.discourse.group/t/double-exports-in-txtexport/58 ??

[wasanti]

Hello. It is similar but not the same, the .txt file generated by the code of task01 gives the concentration values at the lowest x value and repeats this line (2*n) -1 times. Here n is the number of vertices specified in the configuration of the simulation.

[RemDelaporteMathurin]

Can you provide the TXT output please

[wasanti]

Here you are: mobile_concentration.txt

[RemDelaporteMathurin]

Ok so the content looks like this:

x,t=1.00e-01s,t=2.00e-01s,t=5.00e-01s,t=1.00e+00s
6.999999999999999816e-09,3.745935295323514400e+16,4.862470564827332000e+16,5.637923701412466400e+16,5.702999233687084800e+16
6.999999999999999816e-09,3.745935295323514400e+16,4.862470564827332000e+16,5.637923701412466400e+16,5.702999233687084800e+16
6.999999999999999816e-09,3.745935295323514400e+16,4.862470564827332000e+16,5.637923701412466400e+16,5.702999233687084800e+16
6.999999999999999816e-09,3.745935295323514400e+16,4.862470564827332000e+16,5.637923701412466400e+16,5.702999233687084800e+16
6.999999999999999816e-09,3.745935295323514400e+16,4.862470564827332000e+16,5.637923701412466400e+16,5.702999233687084800e+16
6.999999999999999816e-09,3.745935295323514400e+16,4.862470564827332000e+16,5.637923701412466400e+16,5.702999233687084800e+16
6.999999999999999816e-09,3.745935295323514400e+16,4.862470564827332000e+16,5.637923701412466400e+16,5.702999233687084800e+16
6.999999999999999816e-09,3.745935295323514400e+16,4.862470564827332000e+16,5.637923701412466400e+16,5.702999233687084800e+16
6.999999999999999816e-09,3.745935295323514400e+16,4.862470564827332000e+16,5.637923701412466400e+16,5.702999233687084800e+16
6.999999999999999816e-09,3.745935295323514400e+16,4.862470564827332000e+16,5.637923701412466400e+16,5.702999233687084800e+16
6.999999999999999816e-09,3.745935295323514400e+16,4.862470564827332000e+16,5.637923701412466400e+16,5.702999233687084800e+16
6.999999999999999816e-09,3.745935295323514400e+16,4.862470564827332000e+16,5.637923701412466400e+16,5.702999233687084800e+16
6.999999999999999816e-09,3.745935295323514400e+16,4.862470564827332000e+16,5.637923701412466400e+16,5.702999233687084800e+16
6.999999999999999816e-09,3.745935295323514400e+16,4.862470564827332000e+16,5.637923701412466400e+16,5.702999233687084800e+16

All the rows have the same x value.

This is the expected output:

x,t=1.00e-01s,t=2.00e-01s,t=5.00e-01s,t=1.00e+00s
6.999999999999999816e-09,3.488893674270494800e+16,4.645798178830062400e+16,5.586287860482966400e+16,5.702014515046599200e+16
0.000000000000000000e+00,9.999999999999956250e+14,9.999999999999956250e+14,9.999999999999998750e+14,9.999999999999998750e+14
1.399999999999999963e-08,6.866589844313862400e+16,9.180385421290123200e+16,1.106135543397368800e+17,1.129280760092114560e+17
6.999999999999999816e-09,3.488893674270494000e+16,4.645798178830064000e+16,5.586287860482966400e+16,5.702014515046600800e+16
2.099999999999999945e-08,1.023311246504487360e+17,1.370377225071364000e+17,1.652520389328216640e+17,1.687237928826510400e+17
1.399999999999999963e-08,6.866589844313862400e+16,9.180385421290121600e+16,1.106135543397368800e+17,1.129280760092114880e+17
2.799999999999999926e-08,1.358848549007206720e+17,1.821596919021333440e+17,2.197783441119523200e+17,2.244072960771930880e+17
2.099999999999999945e-08,1.023311246504487360e+17,1.370377225071364000e+17,1.652520389328216640e+17,1.687237928826510720e+17
3.499999999999999577e-08,1.693273287107725440e+17,2.271698676257088320e+17,2.741924816046502080e+17,2.799785858992372800e+17
2.799999999999999926e-08,1.358848549007206720e+17,1.821596919021333120e+17,2.197783441119523520e+17,2.244072960771931200e+17

I haven't been able to reproduce your issue on Ubuntu and the latest fenics version. Have you modified the task in any way?

[wasanti]

No, I pulled the git repository and made sure to that git diff gives an empty result.

How would I print the fenics version that FESTIM accesses?

[RemDelaporteMathurin]

print(fenics.version) I would say

[wasanti]

Ok then it is: 2019.2.0.dev0

[RemDelaporteMathurin]

This is what I have when I run conda list fenics

# Name                    Version                   Build  Channel
fenics                    2019.1.0            h8783aee_48    conda-forge
fenics-dijitso            2019.1.0          pyhd8ed1ab_39    conda-forge
fenics-dolfin             2019.1.0        py311h57ea513_48    conda-forge
fenics-ffc                2019.1.0          pyhd8ed1ab_39    conda-forge
fenics-fiat               2019.1.0          pyhd8ed1ab_39    conda-forge
fenics-libdolfin          2019.1.0            h894a989_48    conda-forge
fenics-ufl                2019.1.0        py311h38be061_38    conda-forge

and

import dolfin as df
print (df.__version__)

gives 2019.1.0

[RemDelaporteMathurin]

@wasanti you seem to have a development version of legacy-fenics. I'm trying to conda install it to try and reproduce your issue.

Can you try and install the 2019.1.0 version?

[wasanti]

It is getting wilder: When I installed "fenics:amd64 (1:2019.1.0.1~ppa1~disco1)" from the fenics repo or "fenics:amd64 (1:2019.2.0.5)" from the debian repo, the "fenix-version" command continues: to yield: "2019.2.0.dev0".

The error also persists for both installs.

[RemDelaporteMathurin]

This sounds like you have conflicting versions. Can you clean up everything (remove fenics completely) and reinstall the correct (non-dev) version?

[wasanti]

When doing this, I get stuck in a dependency/version clash between fenics, dolfin and festim. I wont rule out that there is a way out but I would leave it at that since the Debian version (bullseye) I am using will soon be replaced. I'll try again when I install a newer version of Debian. I further tested installing via conda and that runs fine.