To creat the samples.json file, feed the script with the full path to the fastq directory:
# conda environment
source activate snakemake
python3 write_sample2_json.py --fastq_dir ~/git_repos/RNA-seq-analysis/RNA-seq-snakemake-pipeline/myfastqs#dry run
snakemake -np
## work flow diagram
snakemake --forceall --dag | dot -Tpng | display
# run
snakemake 
Fold structure for the fastq files is like this:
myfastqs/
├── sample1
│ ├── sample1_R1.fastq.gz
│ └── sample1_R2.fastq.gz
├── sample2
│ ├── sample2_R1.fastq.gz
│ └── sample2_R2.fastq.gz
└── sample3
├── sample3_R1.fastq.gz
└── sample3_R2.fastq.gz