Methods

Reading and accessing data

Data

All methods except generate_3d_image are methods applied to the Data object.

Data(path, id_type='numeric', id_list=None)

Attributes:

• path: str
  • Path to MRICloud data text file
• id_type: str, {'numeric', 'filename', 'custom'}, default = 'numeric'
  • Type of subject ID formatting
• id_list: list, default = None
  • List of custom subject IDs
• df: DataFrame
  • DataFrame generated from path

---

get_data

Retrieve DataFrame of a given data object.

get_data()

Parameters:

• None

Returns: DataFrame

---

get_id

Retrieve list of unique subject IDs.

get_id()

Parameters:

• None

Returns: Series

---

Manipulating data

rename_subject

Rename a specific subject ID.

rename_subject(old, new)

Parameters:

• old: str
  • Old subject name to be replaced
• new: str
  • New subject name

Returns: DataFrame

long_to_wide

Convert default long form data to a wide format.

long_to_wide()

Parameters:

• None

Returns: DataFrame

---

Visualization

generate_sunburst

Generate a Plotly Express sunburst Figure model.

generate_sunburst(type, id, base_level=5)

Parameters:

• type: int, {1, 2}
  • Type of hierarchical view
• id: str
  • Subject ID
• base_level: int, {1, 2, 3, 4, 5}, default = 5
  • Lowest hierarchical level to include

Returns: Figure

---

generate_treemap

Generate a Plotly Express treemap Figure model.

generate_treemap(type, id, base_level=5)

Parameters:

• type: int, {1, 2}
  • Type of hierarchical view
• id: str
  • Subject ID
• base_level: int, {1, 2, 3, 4, 5}, default = 5
  • Lowest hierarchical level to include

Returns: Figure

---

generate_icicle

Generate a Plotly Express icicle Figure model.

generate_icicle(type, id, base_level=5)

Parameters:

• type: int, {1, 2}
  • Type of hierarchical view
• id: str
  • Subject ID
• base_level: int, {1, 2, 3, 4, 5}, default = 5
  • Lowest hierarchical level to include

Returns: Figure

---

generate_bar

Generates a Plotly Express bar graph Figure.

generate_bar(type, level, id, x='ID', y='Prop', log_y=False)

Parameters:

• type: int, {1, 2}
  • Type of hierarchical view
• level: int, {1, 2, 3, 4, 5}
  • Hierarchical level of interest
• id: list, default = None
  • Subjects of interest
• x: str, {'ID', 'Object'}, default = 'ID'
  • Independent variable
• y: str, {'Prop', 'Volume'}, default = 'Prop'
  • Dependent variable
• log_y: bool, default = False
  • Logarithm of dependent variable

Returns: Figure

---

generate_mean_diff

Generate a Plotly Express mean difference plot Figure.

generate_mean_diff(type, level, color='ID', id=None)

Parameters:

• type: int, {1, 2}
  • Type of hierarchical view
• level: int, {1, 2, 3, 4, 5}
  • Hierarchical level of interest
• color: str, {'ID', 'Object'}, default = 'ID'
  • Variable to organize data by color
• id: list, default = None
  • Subjects of interest

Returns: Figure

---

generate_corr_matrix

Generate a Plotly Express heatmap Figure of a correlation matrix.

generate_corr_matrix(type, level, id=None)

Parameters:

• type: int, {1, 2}
  • Type of hierarchical view
• level: int, {1, 2, 3, 4, 5}
  • Hierarchical level of interest
• id: list, default = None
  • Subjects of interest

Returns: Figure

---

Modeling and covariate analysis

append_covariate_data

Append covariate dataset to data object.

append_covariate_data(path, icv=False, tbv=False)

Parameters:

• path: str
  • Path to covariate dataset file
• icv: bool, default = False
  • Append intracranial volume to covariate dataset
• tbv: bool, default = False
  • Append total brain volume to covariate dataset

Returns: DataFrame

---

normalize_covariate_data

Normalize region data in covariate dataset by ICV, TBV, or ICV + TBV.

normalize_covariate_data(covariate_dataset, normalizing_factor)

Parameters:

• covariate_dataset: DataFrame
  • Covariate dataset to be normalized
• normalizing_factor: str, {'icv, tbv, icv_tbv'}
  • Variable to normalize region volumes by

Returns: DataFrame

---

OLS

Run statsmodels Ordinary Least Squares regression on data object.

OLS(covariate_dataset, covariates, outcome, log=False, residual_plot=False)

Parameters:

• covariate_dataset: DataFrame
  • Dataset containing the covariates and outcome
• covariates: list
  • Covariates to include in analysis (x, independent covariates)
• outcome: str
  • Outcome of interest (y, dependent covariate)
• log: bool, default = False
  • Logaritm of covariates
• residual_plot: bool, default = False
  • Return a residual plot of analysis results as Plotly Figure

Returns: RegressionResultsWrapper.summary(), Figure

---

Logit

Run statsmodels Logit regression on data object.

Logit(covariate_dataset, covariates, outcome, log=False, roc_plot=False)

Parameters:

• covariate_dataset: DataFrame
  • Dataset containing the covariates and outcome
• covariates: list
  • Covariates to include in analysis (x, independent covariates)
• outcome: str
  • Outcome of interest (y, dependent covariate)
• log: bool, default = False
  • Logaritm of covariates
• roc_plot: bool, default = False
  • Return a residual plot of analysis results as Plotly Figure

Returns: RegressionResultsWrapper.summary(), Figure

---

Imaging

Generates a subplot or single image (if 'nrows' and 'ncols' is 1) of region-specific brain images on a template brain

generate_3d_image

generate_3d_image(img_path, regions, view, nrows, ncols, slice_n=0)

Parameters:

• img_path: str
• regions: list
• view: int, {0 (horizontal), 1 (sagittal), 2 (coronal)}
• nrows: int, {1, 2, 3, 4, 5, 6, 7}
• ncols: int, {1, 2, 3, 4, 5, 6, 7}
• slice_n: int, default = 0