I face a strange problem with no answer found : I want to represent this dataframe (which is a time series) : ts_velo.head() type_velo nombre date 2016-09-01 VAE 0.980769 2016-09-01 Vélos 11.865385 2016-10-01 VAE 0.826087 2016-10-01 Vélos 9.152174 2016-11-01 VAE 0.711538 (...) I can draw this graph : sns.lineplot(x=ts_velo.index, y='nombre', data=ts_velo, hue='type_velo') But I don't achieve to use dashed lines instead of solid lines. I tried several things I read on stackoverflow, for example : sns.lineplot(x=ts_velo.index, y='nombre', data=ts_velo, hue='type_velo', dashes=[(2, 2), (2, 2)]) sns.lineplot(x=ts_velo.index, y='nombre', data=ts_velo, hue='type_velo', dashes=True) sns.lineplot(x=ts_velo.index, y='nombre', data=ts_velo, hue='type_velo', dashes='--') sns.lineplot(x=ts_velo.index, y='nombre', data=ts_velo, hue='type_velo', ls='--') g = sns.lineplot(x=ts_velo.index, y='nombre', data=ts_velo, hue='type_velo') g.lines[0].set_linestyle("--") But nothing works. An idea of why I get this problem ? Continue reading...