From a129253b66eb6e8c5655fe7a458f6f1a33d8bca0 Mon Sep 17 00:00:00 2001
From: oboulle <olivier.boulle@inria.fr>
Date: Thu, 2 May 2024 11:38:09 +0200
Subject: [PATCH] init from csv method
---
read_matrix.py | 106 +++++++++++++++++++++++++++++++++++++++++--------
1 file changed, 89 insertions(+), 17 deletions(-)
diff --git a/read_matrix.py b/read_matrix.py
index 5ec0b0f..d1ee270 100755
--- a/read_matrix.py
+++ b/read_matrix.py
@@ -18,7 +18,7 @@ class Minimizer_matrix:
class to handle matrix stuff
"""
- def __init__(self, reads_path: str, minimizers_dict: dict) -> None:
+ def __init__(self) -> None:
self.lines = [] # list of lines
self.soluce_dict = {} # cluster_name : list of reads from the cluster
@@ -26,28 +26,29 @@ class Minimizer_matrix:
self.line_number = 0
self.column_number = 0
- self.generate_matrix(reads_path, minimizers_dict)
-
-
- def generate_matrix(self, reads_path: str, minimizers_dict: dict) -> None:
+ @classmethod
+ def init_from_reads(cls, reads_path: str, minimizers_dict: dict):
"""
- fill the matrix from a .fastq of reads and a dict of minimizers from these reads
+ init the matrix object from a .fastq of reads and a dict of minimizers from these reads
"""
+ # create the object
+ new_matrix = Minimizer_matrix()
+
# read the .fastq file and get basic info
reads_dict = dfr.read_fastq(reads_path)
reads_name_list = list(reads_dict.keys())
- self.line_number = len(minimizers_dict)
- self.column_number = len(reads_dict)
+ new_matrix.line_number = len(minimizers_dict)
+ new_matrix.column_number = len(reads_dict)
# shuffle the reads and save the original clusters in a dict
random.shuffle(reads_name_list)
for i, read_name in enumerate(reads_name_list):
read_cluster = int(read_name.split("_")[1])
- self.soluce_dict[read_cluster] = self.soluce_dict.get(read_cluster, []) + ["r"+str(i)]
+ new_matrix.soluce_dict[read_cluster] = new_matrix.soluce_dict.get(read_cluster, []) + ["r"+str(i)]
- self.soluce_dict = dict(sorted(self.soluce_dict.items()))
+ new_matrix.soluce_dict = dict(sorted(new_matrix.soluce_dict.items()))
# construction of the matrix
@@ -59,14 +60,83 @@ class Minimizer_matrix:
# test if minimizer is in the read
if minimizer in read_seq or minimizer in read_seq_rv:
minimizer_line.append("1")
- self.ones_counter += 1
+ new_matrix.ones_counter += 1
else:
minimizer_line.append("0")
- self.lines.append(minimizer_line)
+ new_matrix.lines.append(minimizer_line)
# show matrix density
- cases_counter = self.line_number*self.column_number
- print(str(self.ones_counter)+ "/"+ str(cases_counter) +" = "+ str(round(100*self.ones_counter/cases_counter, 1))+"%")
+ cases_counter = new_matrix.line_number*new_matrix.column_number
+ print(str(new_matrix.ones_counter)+ "/"+ str(cases_counter) +" = "+ str(round(100*new_matrix.ones_counter/cases_counter, 1))+"%")
+
+ return new_matrix
+
+
+ @classmethod
+ def init_from_matrix_csv(cls, matrix_csv_path, soluce_path=None):
+ """
+ init the matrix object from a csv file #TODO optional soluce file
+ """
+ # create the object
+ new_matrix = Minimizer_matrix()
+
+ with open(matrix_csv_path, 'r') as input_csv:
+ input_csv.readline() # skip columns names
+ line = input_csv.readline()
+ while line != "":
+ matrix_row = line.replace("\n","").split(",")[1:]
+ new_matrix.lines.append(matrix_row)
+ new_matrix.ones_counter += sum(int(k) for k in matrix_row)
+ line = input_csv.readline()
+ new_matrix.line_number = len(new_matrix.lines)
+ new_matrix.column_number = len(new_matrix.lines[0])
+
+ if soluce_path is not None:
+ # load the soluce dict from a file
+ with open(soluce_path, 'r') as input_soluce:
+ line = input_soluce.readline()
+ while line != "":
+ cluster_name = int(line.split(":")[0])
+ cluster_columns = line.split(":")[1].replace("\n","").split(",")
+ new_matrix.soluce_dict[cluster_name] = cluster_columns
+ line = input_soluce.readline()
+
+ return new_matrix
+
+
+ def get_representative(self):
+ repr_list = []
+
+ def prod_scalaire(line1, line2):
+ for k in range(self.column_number):
+ if line1[k] == "1" and line2[k] == "1":
+ return True
+ return False
+
+ repr_list.append(self.lines[0])
+
+ for i in range(5):
+ compatible_new_repr = False
+ while not compatible_new_repr:
+ compatible_new_repr = True
+ repr = random.choice(self.lines)
+
+ for other_repr in repr_list:
+ if prod_scalaire(repr, other_repr):
+ compatible_new_repr = False
+ break
+ print(i+1,"/5",)
+ repr_list.append(repr)
+
+ with open("test_repr", 'w') as output_file:
+ for repr in repr_list:
+
+ output_file.write(str(repr)+"\n")
+
+ print(repr_list)
+
+
+
def print_soluce(self, solutions_path: str) -> None:
@@ -257,7 +327,7 @@ def matrix_generation(column_number: int, dir_path: str) -> None:
minimizer_dict = get_minimizers(reads_path)
- matrix_test = Minimizer_matrix(reads_path, minimizer_dict) # init a matrix class object
+ matrix_test = Minimizer_matrix.init_from_reads(reads_path, minimizer_dict) # init a matrix class object from the reads file & minimizers
# save the matrix and solution
matrix_test.print_soluce(solutions_path)
@@ -308,11 +378,13 @@ if __name__ == "__main__":
print("generate matrix...")
- column_number = 1000
+ column_number = 100
dir_path = "matrix_tests/matrix_100_test/"
- matrix_generation(column_number, dir_path)
+ dir_path = "matrix_tests/matrix_10k/"
+ matrix = Minimizer_matrix.init_from_matrix_csv(dir_path+"matrix_10k.csv", dir_path+"soluce.txt")
+ matrix.get_representative()
#display_results(dir_path+"/matrix_100/matrix.csv", dir_path+"/matrix_100_results", dir_path+"/matrix_100/matrix_100_ordered_results.csv")
print("\tcompleted !")
--
GitLab