Writing tabular data

With the write_tabular app, cogent3 “TabularTypes” (Table, DictArray, DistanceMatrix) are supported for writing to disk.

Let’s generate a cogent3 Table to use in the examples below. One way to do that is by applying the tabulate_stats app to a model result.

from cogent3 import get_app

# load alignment
load_aligned_app = get_app("load_aligned", moltype="dna")
aln = load_aligned_app("data/primate_brca1.fasta")

# fit GN model
gn_model_app = get_app("model", "GN", tree="data/primate_brca1.tree")
model_result = gn_model_app(aln)

# tabulate the model result
tabulator = get_app("tabulate_stats")
model_result_tab = tabulator(model_result)
motif_params = model_result_tab["motif params"]
print(type(motif_params))
motif_params

<class 'cogent3.util.table.Table'>

motif params

A	C	G	T
0.3757	0.1742	0.2095	0.2406

1 rows x 4 columns

Writing a CSV file

To write in CSV format, we create the write_tabular app with format="csv".

from cogent3 import get_app, open_data_store

out_dstore = open_data_store(path_to_dir, mode="w", suffix="csv")

write_tabular_app = get_app("write_tabular", data_store=out_dstore, format="csv")
write_tabular_app(motif_params, identifier="gn_model_results.csv")

DataMember(data_store=/home/runner/work/cogent3.github.io/cogent3.github.io/c3org/doc/doc/tmp1kx89c5y, unique_id=gn_model_results.csv)

Writing a TSV file

To write in TSV format, we create the write_tabular app with format="tsv".

from cogent3 import get_app, open_data_store

out_dstore = open_data_store(path_to_dir, mode="w", suffix="tsv")

write_tabular_app = get_app("write_tabular", data_store=out_dstore, format="tsv")
write_tabular_app(motif_params, identifier="gn_model_results.tsv")

DataMember(data_store=/home/runner/work/cogent3.github.io/cogent3.github.io/c3org/doc/doc/tmp065_6kyn, unique_id=gn_model_results.tsv)

Using write_tabular in a composed process

Instead of applying the apps sequentially as above, we can add apps into a composed process, and apply the process to a data store. In this example, we define a process that calculates an unaligned distance measure between sequences, writing these estimated distances to a tsv file.

from cogent3 import get_app, open_data_store

loader = get_app("load_unaligned", moltype="dna")
jdist = get_app("jaccard_dist")
out_dstore = open_data_store(path_to_dir, mode="w", suffix="tsv")
writer = get_app("write_tabular", data_store=out_dstore, format="tsv")

process = loader + jdist + writer

in_dstore = open_data_store("data", suffix="fasta", mode="r", limit=2)

result = process.apply_to(in_dstore)
result.describe

Directory datastore

record type	number
completed	2
not_completed	0
logs	1

3 rows x 2 columns

Tip

When running this code on your machine, remember to replace path_to_dir with an actual directory path.