Snakemake执行ph_final/q2r报错：目标规则不可含通配符

执行的命令 #

• snakemake -j8 ph_final
• snakemake -j8 q2r

报错信息 #

Building DAG of jobs...
WorkflowError:
Target rules may not contain wildcards. Please specify concrete files or a rule without wildcards at the command line, or have a rule without wildcards at the very top of your workflow (e.g. the typical "rule all" which just collects all results you want to generate in the end).

使用的Snakemake文件 #

configfile: "config.yml"

rule ph_target:
    input:
        expand("ph/V{volume}/ph.in",
            volume=config["eos"]["volumes"]),
        expand("ph/V{volume}/scf.in",
            volume=config["eos"]["volumes"])

rule ph_in:
    input:
        structure="{workdir}/init.qe",
        template_scf=config["phonon"]["template"]["scf"],
        template_ph=config["phonon"]["template"]["ph"],
        job_sh=config["phonon"]["job_sh"]
    output:
        scf_in="{workdir}/scf.in",
        ph_in="{workdir}/ph.in",
        job_sh="{workdir}/job.sh"
    shell:
        '''
        cat {input.template_scf} > {output.scf_in}
        cat {input.structure} >> {output.scf_in}
        cp {input.template_ph} {output.ph_in}
        cp {input.job_sh} {output.job_sh}
        '''

rule ph_structures:
    input:
        structure="relax/V{volume}/final.qe"
    output:
        structure="ph/V{volume}/init.qe"
    shell:
        '''
        cp {input.structure} {output.structure}
        '''

rule ph_final:
    input:
        scf_out="{workdir}/scf.out",
        ph_out="{workdir}/ph.out"
    output:
        scf_done="{workdir}/scf.out.done",
        ph_done="{workdir}/ph.out.done",
        final="{workdir}/ph.final"
    shell:
        '''
        bash scripts/check-done.sh {input.scf_out}
        bash scripts/check-done.sh {input.ph_out}
        bash scripts/check-ph-final.sh {wildcards.workdir}
        '''

rule q2r:
    input:
        ph_final="{workdir}/ph.final",
        template="templates/q2r.in"
    output:
        input="{workdir}/q2r.in",
        out="{workdir}/q2r.out",
        err="{workdir}/q2r.err",
        done="{workdir}/q2r.out.done",
        fc=expand("{{workdir}}/{prefix}.fc", prefix=config["system"]["prefix"])
    shell:
        '''
        cp {input.template} {output.input}
        bash scripts/q2r.sh {output.input}
        bash scripts/check-done.sh {output.out}
        '''

rule matdyn_qha:
    input:
        ph_final="{workdir}/ph.final",
        template="templates/matdyn-qha.in",
        fc=expand("{{workdir}}/{prefix}.fc", prefix=config["system"]["prefix"]),
        weights="weights/weights.txt"
    output:
        input="{workdir}/matdyn-qha.in",
        out="{workdir}/matdyn-qha.out",
        err="{workdir}/matdyn-qha.err",
        done="{workdir}/matdyn-qha.out.done",
        freq=expand("{{workdir}}/{prefix}.freq", prefix=config["system"]["prefix"]),
        eig=expand("{{workdir}}/{prefix}.eig", prefix=config["system"]["prefix"]),
        vec=expand("{{workdir}}/{prefix}.vec", prefix=config["system"]["prefix"])
    shell:
        '''
        cp {input.template} {output.input}
        echo "`cat {input.weights}`" | wc -l >> {output.input}
        cat {input.weights} >> {output.input}
        bash scripts/matdyn.sh {output.input}
        bash scripts/check-done.sh {output.out}
        '''

rule input01:
    params:
        comment=config["system"]["prefix"]
    input:
        expand("ph/V{volume}/ph.final", volume=config["eos"]["volumes"]),
        expand("ph/V{volume}/ph.out.done", volume=config["eos"]["volumes"]),
        freq=expand("ph/V{volume}/{prefix}.freq", prefix=config["system"]["prefix"], volume=config["eos"]["volumes"]),
        weights="weights/weights.txt"
    output:
        input01="input01"
    shell:
        '''
        python3 scripts/make-input01.py -c {params.comment} {input.freq} > {output.input01}
        echo "" >> {output.input01}
        echo "weights" >> {output.input01}
        cat {input.weights} >> {output.input01}
        '''

rule input01_ordered:
    params:
        prefix=config["system"]["prefix"],
        comment=config["system"]["prefix"]
    input:
        expand("ph/V{volume}/ph.final", volume=config["eos"]["volumes"]),
        expand("ph/V{volume}/ph.out.done", volume=config["eos"]["volumes"]),
        freq=expand("ph/V{volume}/{prefix}.freq", prefix=config["system"]["prefix"], volume=config["eos"]["volumes"]),
        eig=expand("ph/V{volume}/{prefix}.eig", prefix=config["system"]["prefix"], volume=config["eos"]["volumes"]),
        weights="weights/weights.txt"
    output:
        input01="input01.ordered"
    shell:
        '''
        python3 scripts/make-input01.py --sort-by {params.prefix}.eig -c {params.comment} {input.freq} > {output.input01}
        echo "" >> {output.input01}
        echo "weights" >> {output.input01}
        cat {input.weights} >> {output.input01}
        '''

rule weights_target:
    params:
        kpts=expand("{nk1},{nk2},{nk3}",
            nk1=config["phonon"]["matdyn_kpts"][0],
            nk2=config["phonon"]["matdyn_kpts"][1],
            nk3=config["phonon"]["matdyn_kpts"][2],
        )
    input:
        scf=expand("ph/V{volume}/scf.in", volume=config["eos"]["volumes"][0])
    output:
        input="weights/weights.in",
        weights="weights/weights.txt"
    shell:
        '''
        python3 scripts/make-weights-input.py {input.scf} --nk {params.kpts} > {output.input}
        bash scripts/gen_coords2.sh {output.input}
        '''

解决方案 #

报错核心原因：ph_final和q2r规则都包含*{workdir}*通配符，Snakemake要求作为命令行目标的规则不能带通配符——它无法自动判断你要生成哪个具体路径下的产物。

方法1：命令行指定具体目标文件 #

直接给出包含实际路径的输出文件（假设config的volumes列表包含100）：

• 运行ph_final对应任务：snakemake -j8 ph/V100/ph.final
• 运行q2r对应任务：snakemake -j8 ph/V100/q2r.out.done

方法2：添加无通配符的顶层规则（推荐） #

在Snakemake文件最开头新增rule all，用expand展开所有需要生成的最终产物，之后直接运行snakemake -j8就能自动处理所有体积的任务：

configfile: "config.yml"

# 新增顶层规则，明确所有最终产物
rule all:
    input:
        # 所有体积的ph_final产物
        expand("ph/V{volume}/ph.final", volume=config["eos"]["volumes"]),
        # 所有体积的q2r产物
        expand("ph/V{volume}/q2r.out.done", volume=config["eos"]["volumes"]),
        # 可选：添加matdyn_qha或input01的产物
        expand("ph/V{volume}/{prefix}.freq", prefix=config["system"]["prefix"], volume=config["eos"]["volumes"]),
        "input01"

方法3：修改规则为固定路径（不推荐） #

如果只需要处理单一固定路径，可把规则中的{workdir}替换成具体路径（比如ph/V100），但这样会失去处理多体积任务的灵活性。

内容的提问来源于stack exchange，提问作者Bill C Oyomo